Initial Linux ZFS GIT Repo

2008-11-20 12:01:55 -08:00 · 2008-11-20 12:01:55 -08:00 · 34dc7c2f25
commit 34dc7c2f25
444 changed files with 187636 additions and 0 deletions
--- a/4
+++ b/4
@ -0,0 +1,4 @@
+Brian Behlendorf <behlendorf1@llnl.gov>,
+Herb Wartens <wartens2@llnl.gov>,
+Jim Garlick <garlick@llnl.gov>,
+Ricardo M. Correia <Ricardo.M.Correia@sun.com>
--- a/114
+++ b/114
@ -0,0 +1,114 @@
+2008-11-19 Brian Behlendorf <behlendorf1@llnl.gov>
+
+	* : Tag zfs-0.4.0
+
+	* : ZFS project migrated from Subversion which leveraged a 
+	quilt based patch stack to Git and a TopGit managed patch
+	stack.  The new method treats all patches as Git branches
+	which can be more easily shared for distributed development.
+	Consult the top level GIT file for detailed information on
+	how to properly develop for this package using Git+TopGit.
+
+2008-11-12 Brian Behlendorf <behlendorf1@llnl.gov>
+
+	* : Tag zfs-0.3.4
+
+	* zfs-07-create-dev-zfs.patch:
+	Ricardo M. Correia <Ricardo.M.Correia@sun.com>
+	- Make libzfs create /dev/zfs if it doesn't exist.
+
+	* zfs-05-check-zvol-size.patch:
+	Ricardo M. Correia <Ricardo.M.Correia@sun.com>
+	- Properly check zvol size under Linux.
+
+	* zfs-04-no-openat-fdopendir.patch:
+	Ricardo M. Correia <Ricardo.M.Correia@sun.com>
+	- Do not use openat() and fdopendir() since they are not available
+	  on older systems.
+
+	* zfs-03-fix-bio-sync.patch:
+	Ricardo M. Correia <Ricardo.M.Correia@sun.com>
+	- Fix memory corruption in RHEL4 due to synchronous IO becoming
+	  asynchronous.
+
+2008-11-06 Brian Behlendorf <behlendorf1@llnl.gov>
+
+	* zfs-02-zpios-fix-stuck-thread-memleak.patch:
+	Ricardo M. Correia <Ricardo.M.Correia@sun.com>
+	- Fix stuck threads and memory leaks when errors occur while writing.
+
+	* zfs-01-zpios-arg-corruption.patch:
+	Ricardo M. Correia <Ricardo.M.Correia@sun.com>
+	- Fix zpios cmd line argument corruption problem.
+
+	* zfs-00-minor-fixes.patch: 
+	Ricardo M. Correia <Ricardo.M.Correia@sun.com>
+	- Minor build system improvements
+	- Minor script improvements
+	- Create a full copy and not a link tree with quilt
+	- KPIOS_MAJOR changed from 231 to 232
+	- BIO_RW_BARRIER flag removed from IO request
+
+2008-06-30 Brian Behlendorf <behlendorf1@llnl.gov>
+
+	* : Tag zfs-0.3.3
+
+	* : Minor script updates and tweaks to be compatible with
+	the latest version of the SPL.
+
+2008-06-13 Brian Behlendorf <behlendorf1@llnl.gov>
+
+	* vdev_disk.diff: Replace vdev_disk implementation which was
+	based on the kmalloc'ed logical address space with a version
+	which works with vmalloc'ed memory in the virtual address space.
+	This was done to support the new SPL slab implementation which
+	is based on virtual addresses to avoid the need for contigeously
+	allocated memory.
+
+2008-06-05 Brian Behlendorf <behlendorf1@llnl.gov>
+
+	* arc-vm-integration.diff: Reduce maximum default arc memory 
+	usage to 1/4 of total system memory.  Because all the bulk data
+	is still allocated on the slab memory fragmentation is a serious
+	concern.  To address this in the short term we simply need to
+	leave lots of free memory.
+
+	* fix-stack.diff: First step towards reducing stack usage so
+	we can run the full ZFS stack using a stock kernel.
+
+2008-06-04 Brian Behlendorf <behlendorf1@llnl.gov>
+
+	* : Tag zfs-0.3.2
+
+	* : Extensive improvements to the build system to detect kernel
+	API changes so we can flexibly build with a wider range of kernel
+	versions.  The code has now been testing with the 2.6.18-32chaos
+	and 2.6.25.3-18.fc9 kernels, however we should also be compatible
+	with other kernels in the range of 2.6.18-2.6.25.  The only
+	remaining issue preventing us from running with a stock 
+	kernel is ZFS stack usage.
+
+2008-05-21 Brian Behlendorf <behlendorf1@llnl.gov>
+
+	* : Tag zfs-0.3.1
+
+	* : License headers including URCL added for release.
+
+2008-05-21 Brian Behlendorf <behlendorf1@llnl.gov>
+
+	* : Tag zfs-0.3.0
+
+	* configure.ac: Improved autotools support and configurable debug.
+
+2008-05-15 Brian Behlendorf <behlendorf1@llnl.gov>
+
+	* : Updating original ZFS sources to build 89 which
+	includes the new write throttling changes plus support
+	for using ZFS as your root device.  Neither of which
+	will work exactly right without some more work but this
+	gets us much closers to the latest source.
+	
+
+2008-02-28 Brian Behlendorf <behlendorf1@llnl.gov>
+
+	* : First attempt based on SPL module and zfs-lustre sources
--- a/22
+++ b/22
@ -0,0 +1,22 @@
+This notice is required to be provided under our contract with the 
+U.S. Department of Energy (DOE).  This work was produced at the 
+Lawrence Livermore National Laboratory under Contract with the DOE.
+
+Neither the United States Government nor the Lawrence Livermore National
+Security, LLC. nor any of their employees, makes any warranty, express
+or implied, or assumes any liability or responsibility for the accuracy,
+completeness, or usefulness of any information, apparatus, product,
+or process disclosed, or represents that its use would not infringe
+privately-owned rights.
+
+Also, reference herein to any specific commercial products, process,
+or services by trade name, trademark, manufacturer or otherwise does
+not necessarily constitute or imply its endorsement, recommendation,
+or favoring by the United States Government or the  Lawrence Livermore
+National Security, LLC.  The views and opinions of authors expressed 
+herein do not necessarily state or reflect those of the United States 
+Government or the  Lawrence Livermore National Security, LLC., and 
+shall not be used for advertising or product endorsement purposes.
+
+The precise terms and conditions for copying, distribution and
+modification are specified in the file "COPYING".
--- a/162
+++ b/162
@ -0,0 +1,162 @@
+=========================== WHY USE GIT+TOPGIT? ==========================
+
+Three major concerns were on our mind when setting up this project.
+
+  o First we needed to structure the project in such a way that it would be
+    easy to rebase all of our changes on the latest official ZFS release
+    from Sun.  We absolutely need to be able to benefit from the upstream
+    improvements and not get locked in to an old version of the code base.
+
+  o Secondly, we wanted to be able to easily manage our changes in terms
+    of a patch stack.  This allows us to easily isolate specific changes
+    and push them upstream for inclusion.  It also allows us to easily
+    update or drop specific changes based on what occurs upstream.
+
+  o Thirdly we needed our DVCS to be integrated with the management of this
+    patch stack.  We have tried other methods in the past such as SVN+Quilt
+    but have found managing the patch stack becomes cumbersome.  By using
+    Git+TopGit to more tightly integrate our patch stack in to the repo
+    we expect several benefits.  One of the most important will be the
+    ability to easily work on the patch stack with a distributed developer
+    team, additionally the repo can track patch history, and we can utilize
+    Git to merge patches and resolve conflicts.
+
+TopGit is designed to specifically address these concerns by providing 
+tools to simplify the handling of large numbers of interdependent topic 
+branches.  When using a TopGit aware repo every topic branch represents
+a 'patch' and that branch references its dependent branches.  The union
+of all these branches is your final source base.
+
+========================= SETTING UP GIT+TOPGIT ==========================
+
+First off you need to install a Git package on your system.  For my
+purposes I have been working on a RHEL5 system with git version 1.5.4.5
+installed and it has been working well.  You will also need to go get
+the latest version of TopGit which likely is not packaged nicely so you
+will need to build it from source.  You can use Git to clone TopGit
+from the official site here and your all set:
+
+        > git clone http://repo.or.cz/w/topgit.git
+        > make
+        > make install    # Default installs to $(HOME)
+
+========================== TOPGIT AND ZFS ================================
+
+One you have Git and TopGit installed you will want to clone a copy of
+the Linux ZFS repo.  While this project does not yet have a public home
+it hopefully will some day.  In the meanwhile if you have VPN access to
+LLNL you can clone the latest official repo here.  Cloning a TopGit
+controlled repo is very similar to cloning a normal Git repo, but you
+need to remember to use 'tg remote' to populate all topic branches.
+
+        > git clone http://eris.llnl.gov/git/zfs.git zfs
+        > cd zfs
+        > tg remote --populate origin
+
+Now that you have the Linux ZFS repo the first thing you will probably 
+want to do is have a look at all the topic branches.  TopGit provides
+a summary command which shows all the branches and a brief summary for
+each branch obtained from the .topmsg files.
+
+        > tg summary
+ 0      t/LAST                          [PATCH] LAST
+        t/feature-commit-cb             [PATCH] zfs commit callbacks
+        t/fix-clock-wrap                [PATCH] fix clock wrap
+        t/fix-dnode-cons                [PATCH] fix dnode constructor
+        ...
+
+By convention all TopGit branches are prefixed with 't/', and the Linux
+ZFS repo also introduces the convention that the top most development 
+branch be called 't/LAST".  This provides a consistent label to be used
+when you need to reference the branch which contains the union of all 
+topic branches.
+
+One thing you may also notice about the 'tg summary' command is it does
+not show the branches in dependent order.  While this project only expresses
+a single dependency per branch TopGit implements dependencies as a DAC just
+like Git.  To see the dependencies you will need to use the --graphviz
+option and pipe the result to dot for display.  The following command while
+long works fairly well for me.  Longer term it would be nice to update this
+option to use a preferred config options stored in the repo if they exist.
+
+        > tg summary --graphviz | dot -Txlib -Nfontsize=8 -Eminlen=0.01 \
+        -Grankdir=LR -Nheight=0.3 -Nwidth=2 -Nfixedsize=true
+
+========================= UPDATING A TOPIC BRANCH ========================
+
+Updating a topic branch in TopGit is a pretty straight forward but there
+are a few rules you need to be aware of.  The basic process involves 
+checking out the relevant topic branch where the changes need to be made,
+making the changes, committing the changes to the branch and then merging
+those changes in to dependent branches.  TopGit provides some tools to make
+this pretty easy, although it may be a little sluggish.  Here is an example:
+
+        > git checkout t/feature-commit-cb  # Checkout the proper branch
+        > ...update branch...               # Update the branch
+        > git commit -a                     # Commit your changes
+        > git checkout t/LAST               # Checkout the LAST branch
+        > tg update                         # Recursively merge in new branch
+
+Assuming you change does not introduce any conflicts your done.  All branches
+were dependent on your change will have had the changed merged in.  If your
+change introduced a conflict you will need to resolve the conflict and then
+continue on with the update.
+
+========================== ADDING A TOPIC BRANCH =========================
+
+Adding a topic branch in TopGit is a little more complicated.  When adding
+a new branch to the end of the patch graph things are pretty easy and TopGit
+does all the work.  However, I expect out common case to be adding patches
+to the middle of the graph.  TopGit will allow you to do this but you must
+be careful to manually update the dependency information in the .topdeps
+file.
+
+        > git co t/existing-topic-branch    # Checkout the branch to add after
+        > tg create t/new-topic-branch      # Create a new topic branch
+        > ...update .topmsg...              # Update the branch message
+        > ...create patch...                # Update with your changes
+        > git commit -a                     # Commit your changes
+        > git co t/dependent-topic-branch   # Checkout dependent branch
+        > ...update .topdeps...             # Manually update dependencies
+        > git commit -a                     # Commit your changes
+        > tg update                         # TopGit update
+        > git checkout t/LAST               # Checkout the LAST branch
+        > tg update                         # Recursively merge in new branch
+
+========================= REMOVING A TOPIC BRANCH ========================
+
+Removing a topic branch in TopGit is also currently not very easy.  To remove
+a dependent branch the basic process is to commit a patch which reverts all
+changes on the branch.  Then that reversion must be merged in to all dependent
+branches, the dependencies manually updated and finally the branch removed.
+If the branch is not empty you will not be able to remove it.
+
+        > git co t/del-topic-branch         # Checkout the branch to delete
+        > tg patch | patch -R -p1           # Revert all branch changes
+        > git commit -a                     # Commit your changes
+        > git checkout t/LAST               # Checkout the LAST branch
+        > tg update                         # Recursively merge revert
+        > git co t/dependent-topic-branch   # Checkout dependent branch
+        > ...update .topdeps...             # Manually update dependencies
+        > git commit -a                     # Commit your changes
+        > tg delete t/del-topic-branch      # Delete empty topic branch
+
+============================ TOPGIT TODO =================================
+
+TopGit is still a young package which seems to be under active development
+by its author.  It provides the minimum set of commands needed but there
+are clearly areas which simply have not yet been implemented.  My short
+list of features includes:
+
+  o 'tg summary --deps', option to display a text version of the topic
+    branch dependency DAC.
+
+  o 'tg depend list', list all topic branch dependencies.
+
+  o 'tg depend delete', cleanly remove a topic branch dependency.
+
+  o 'tg create', cleanly insert a topic branch in the middle
+    of the graph and properly take care updating all dependencies.
+
+  o 'tg delete', cleanly delete a topic branch in the middle
+    of the graph and properly take care updating all dependencies.
--- a/6
+++ b/6
@ -0,0 +1,6 @@
+Meta:         1
+Name:         zfs
+Branch:       1.0
+Version:      0.3.4
+Release:      1
+Release-Tags: relext
--- a/Makefile.am
+++ b/Makefile.am
@ -0,0 +1,25 @@
+AUTOMAKE_OPTIONS = foreign dist-zip
+
+SUBDIRS = doc scripts $(BUILDDIR)
+CONFIG_CLEAN_FILES =  aclocal.m4 config.guess config.sub
+CONFIG_CLEAN_FILES += depcomp missing mkinstalldirs
+EXTRA_DIST = autogen.sh
+
+.PHONY: quilt
+quilt:   .quilt-$(BUILDDIR)
+autogen: .autogen-$(BUILDDIR)
+config:  .config-$(BUILDDIR)
+.quilt-$(BUILDDIR):
+	./scripts/quilt.sh -p $(NAME) -b $(BUILDDIR) -s $(SERIESFILE) -d $(PATCHDIR)
+	echo $(BUILDDIR) >$@
+
+unquilt:
+	rm -rf $(BUILDDIR)
+	rm -f .quilt-$(BUILDDIR)
+
+clean-generic:
+
+distclean: unquilt
+
+rpms: dist Makefile
+	rpmbuild -ta $(distdir).tar.gz
--- a/74
+++ b/74
@ -0,0 +1,74 @@
+============================ ZFS KERNEL BUILD ============================
+
+1) Build the SPL (Solaris Porting Layer) module which is designed to
+   provide many Solaris APIs in the Linux kernel which are needed
+   by ZFS.  To build the SPL:
+
+        tar -xzf spl-x.y.z.tgz
+        cd spl-x.y.z
+        ./configure --with-linux=<kernel src>
+        make
+        make check <as root>
+
+2) Build ZFS, this port is based on build 89 of ZFS from OpenSolaris.
+   You will need to have both the kernel and SPL source available.
+   To build ZFS for use as a Linux kernel module (default):
+
+        tar -xzf zfs-x.y.z.tgz
+        cd zfs-x.y.z
+        ./configure --with-linux=<kernel src> \
+                    --with-spl=<spl src>
+        make
+        make check <as root>
+
+========================= ZFS USER LIBRARY BUILD =========================
+
+1) Build ZFS, this port is based on build 89 of ZFS from OpenSolaris.
+   To build ZFS as a userspace library:
+
+        tar -xzf zfs-x.y.z.tgz
+        cd zfs-x.y.z
+        ./configure --zfsconfig=user
+        make
+        make check <as root>
+
+============================ ZFS LUSTRE BUILD ============================
+
+1) Build the SPL (Solaris Porting Layer) module which is designed to
+   provide many Solaris APIs in the Linux kernel which are needed
+   by ZFS.  To build the SPL:
+
+        tar -xzf spl-x.y.z.tgz
+        cd spl-x.y.z
+        ./configure --with-linux=<kernel src>
+        make
+        make check <as root>
+
+2) Build ZFS, this port is based on build 89 of ZFS from OpenSolaris.
+   To build ZFS as a userspace library for use by a Lustre filesystem:
+
+        tar -xzf zfs-x.y.z.tgz
+        cd zfs-x.y.z
+        ./configure --zfsconfig=lustre \
+                    --with-linux=<kernel src> \
+                    --with-spl=<spl src>
+        make
+        make check <as root>
+
+3) Provided is an in-kernel test application called kpios which can be
+   used to simulate a Lustre IO load.  It may be used as a stress test
+   or as a performance to measure your configuration.  To simplify testing
+   there are scripts provided in the scripts/ directory.  A single test
+   can be run as follows:
+
+        WARNING: You MUST update DEVICES in the create-zpool.sh script
+                 to reference the devices you wish to use.
+
+       cd scripts
+       ./load-zfs.sh		# Load the ZFS/SPL module stack
+       ./create-zpool.sh	# Modify DEVICES to list your zpool devices
+       ./zpios.sh		# Modify for your particular kpios test
+       ./unload-zfs.sh		# Unload the ZFS/SPL module stack
+
+Enjoy,
+Brian Behlendorf <behlendorf1@llnl.gov>
--- a/30
+++ b/30
@ -0,0 +1,30 @@
+* We may need a libefi replacement.  It appears libefi is used
+  to determine if the device passed to zpool is a 'whole device'
+  or just a partition of a device.  In the short term I think we
+  can simply treat everything as a partition and be alright.
+
+* We also do not have support for getting Solaris style device
+  ids which is done when a zpool is setup.  We may or may not
+  be able to live without this, the jury is still out.
+
+-----------------------------------------------------------------------
+
+* Port zvol (ZFS volume interface).
+
+* Port zpl (ZFS posix interface).
+
+* Port lustre fsfilt interface to use DMU.
+
+* Andreas issue #1:
+  "the maximum allocation DMU "blocksize" was 128kB and it would be better
+  to be able to get 1MB contiguous allocations for best performance"
+
+* Andreas issue #2:
+  "there would need to be some work done to allow multiple operations to
+  be atomic.  This is needed by Lustre for object creation + LAST_ID
+  updates, unlink + llog updates, etc.  Conceptually this isn't very
+  much work for a phase tree, but I've never looked at the ZFS code."
+
+* Design and implement mechanism for viewing and modifying OST content
+  from user space (by manipulating datasets/objects), possibly
+  by implementing scaled down file system interface.
--- a/autoconf/Makefile.am
+++ b/autoconf/Makefile.am
@ -0,0 +1 @@
+EXTRA_DIST = zfs-build.m4
--- a/autoconf/zfs-build.m4
+++ b/autoconf/zfs-build.m4
@ -0,0 +1,378 @@
+AC_DEFUN([ZFS_AC_CONFIG], [
+
+	AC_ARG_WITH([zfs-config],
+		AS_HELP_STRING([--with-config=CONFIG],
+		[Config file 'kernel|user|lustre']),
+		[zfsconfig="$withval"])
+
+	AC_MSG_CHECKING([zfs config file])
+	if test -z "$zfsconfig" || test ! -r configs/$zfsconfig; then
+		AC_MSG_RESULT([no])
+		AC_MSG_ERROR([
+	        *** Please specify one of the valid config files located
+	        *** in ./configs/ with the '--with-zfs-config=CONFIG' option])
+	fi
+
+	AC_MSG_RESULT([$zfsconfig]);
+	. ./configs/$zfsconfig
+
+	TOPDIR=`/bin/pwd`
+	ZFSDIR=${TOPDIR}/$BUILDDIR
+	LIBDIR=$ZFSDIR/lib
+	CMDDIR=$ZFSDIR/zcmd
+
+	AC_SUBST(TOPDIR)
+	AC_SUBST(ZFSDIR)
+	AC_SUBST(LIBDIR)
+	AC_SUBST(CMDDIR)
+])
+
+AC_DEFUN([ZFS_AC_KERNEL], [
+	ver=`uname -r`
+
+	AC_ARG_WITH([linux],
+		AS_HELP_STRING([--with-linux=PATH],
+		[Path to kernel source]),
+		[kernelsrc="$withval"; kernelbuild="$withval"])
+
+	AC_ARG_WITH(linux-obj,
+		AS_HELP_STRING([--with-linux-obj=PATH],
+		[Path to kernel build objects]),
+		[kernelbuild="$withval"])
+
+	AC_MSG_CHECKING([kernel source directory])
+	if test -z "$kernelsrc"; then
+		kernelbuild=
+		sourcelink=/lib/modules/${ver}/source
+		buildlink=/lib/modules/${ver}/build
+
+		if test -e $sourcelink; then
+			kernelsrc=`(cd $sourcelink; /bin/pwd)`
+		fi
+		if test -e $buildlink; then
+			kernelbuild=`(cd $buildlink; /bin/pwd)`
+		fi
+		if test -z "$kernelsrc"; then
+			kernelsrc=$kernelbuild
+		fi
+		if test -z "$kernelsrc" -o -z "$kernelbuild"; then
+			AC_MSG_RESULT([Not found])
+			AC_MSG_ERROR([
+			*** Please specify the location of the kernel source
+			*** with the '--with-linux=PATH' option])
+		fi
+	fi
+
+	AC_MSG_RESULT([$kernelsrc])
+	AC_MSG_CHECKING([kernel build directory])
+	AC_MSG_RESULT([$kernelbuild])
+
+	AC_MSG_CHECKING([kernel source version])
+	if test -r $kernelbuild/include/linux/version.h && 
+		fgrep -q UTS_RELEASE $kernelbuild/include/linux/version.h; then
+
+		kernsrcver=`(echo "#include <linux/version.h>"; 
+		             echo "kernsrcver=UTS_RELEASE") | 
+		             cpp -I $kernelbuild/include |
+		             grep "^kernsrcver=" | cut -d \" -f 2`
+
+	elif test -r $kernelbuild/include/linux/utsrelease.h && 
+		fgrep -q UTS_RELEASE $kernelbuild/include/linux/utsrelease.h; then
+
+		kernsrcver=`(echo "#include <linux/utsrelease.h>"; 
+		             echo "kernsrcver=UTS_RELEASE") | 
+	        	     cpp -I $kernelbuild/include |
+		             grep "^kernsrcver=" | cut -d \" -f 2`
+	fi
+
+	if test -z "$kernsrcver"; then
+		AC_MSG_RESULT([Not found])
+		AC_MSG_ERROR([
+		*** Cannot determine the version of the linux kernel source.
+		*** Please prepare the kernel before running this script])
+	fi
+
+	AC_MSG_RESULT([$kernsrcver])
+
+	kmoduledir=${INSTALL_MOD_PATH}/lib/modules/$kernsrcver
+        LINUX=${kernelsrc}
+        LINUX_OBJ=${kernelbuild}
+
+	AC_SUBST(LINUX)
+	AC_SUBST(LINUX_OBJ)
+	AC_SUBST(kmoduledir)
+])
+
+AC_DEFUN([ZFS_AC_SPL], [
+
+	AC_ARG_WITH([spl],
+		AS_HELP_STRING([--with-spl=PATH],
+		[Path to spl source]),
+		[splsrc="$withval"; splbuild="$withval"])
+
+	AC_ARG_WITH([spl-obj],
+		AS_HELP_STRING([--with-spl-obj=PATH],
+		[Path to spl build objects]),
+		[splbuild="$withval"])
+
+
+	AC_MSG_CHECKING([spl source directory])
+	if test -z "$splsrc"; then
+		splbuild=
+		sourcelink=/tmp/`whoami`/spl
+		buildlink=/tmp/`whoami`/spl
+
+		if test -e $sourcelink; then
+			splsrc=`(cd $sourcelink; /bin/pwd)`
+		fi
+		if test -e $buildlink; then
+			splbuild=`(cd $buildlink; /bin/pwd)`
+		fi
+		if test -z "$splsrc"; then
+			splsrc=$splbuild
+		fi
+	fi
+
+	if test -z "$splsrc" -o -z "$splbuild"; then
+		sourcelink=/lib/modules/${ver}/source
+		buildlink=/lib/modules/${ver}/build
+
+		if test -e $sourcelink; then
+			splsrc=`(cd $sourcelink; /bin/pwd)`
+		fi
+		if test -e $buildlink; then
+			splbuild=`(cd $buildlink; /bin/pwd)`
+		fi
+		if test -z "$splsrc"; then
+			splsrc=$splbuild
+		fi
+		if test -z "$splsrc" -o -z "$splbuild"; then
+			AC_MSG_RESULT([Not found])
+			AC_MSG_ERROR([
+			*** Please specify the location of the spl source
+			*** with the '--with-spl=PATH' option])
+		fi
+	fi
+
+	AC_MSG_RESULT([$splsrc])
+	AC_MSG_CHECKING([spl build directory])
+	AC_MSG_RESULT([$splbuild])
+
+	AC_MSG_CHECKING([spl source version])
+	if test -r $splbuild/spl_config.h && 
+		fgrep -q VERSION $splbuild/spl_config.h; then
+
+		splsrcver=`(echo "#include <spl_config.h>"; 
+		            echo "splsrcver=VERSION") | 
+		            cpp -I $splbuild |
+	        	    grep "^splsrcver=" | cut -d \" -f 2`
+	fi
+
+	if test -z "$splsrcver"; then
+		AC_MSG_RESULT([Not found])
+		AC_MSG_ERROR([
+		*** Cannot determine the version of the spl source. 
+		*** Please prepare the spl source before running this script])
+	fi
+
+	AC_MSG_RESULT([$splsrcver])
+
+	AC_MSG_CHECKING([spl Module.symvers])
+	if test -r $splbuild/modules/Module.symvers; then
+		splsymvers=$splbuild/modules/Module.symvers
+	elif test -r $kernelbuild/Module.symvers; then
+		splsymvers=$kernelbuild/Module.symvers
+	fi
+
+	if test -z "$splsymvers"; then
+	        AC_MSG_RESULT([Not found])
+	        AC_MSG_ERROR([
+                *** Cannot find extra Module.symvers in the spl source.
+                *** Please prepare the spl source before running this script])
+	fi
+
+	AC_MSG_RESULT([$splsymvers])
+	AC_SUBST(splsrc)
+	AC_SUBST(splsymvers)
+])
+
+AC_DEFUN([ZFS_AC_LICENSE], [
+        AC_MSG_CHECKING([license])
+        AC_MSG_RESULT([CDDL])
+dnl #        AC_DEFINE([HAVE_GPL_ONLY_SYMBOLS], [1],
+dnl #                [Define to 1 if module is licensed under the GPL])
+])
+
+AC_DEFUN([ZFS_AC_DEBUG], [
+        AC_MSG_CHECKING([whether debugging is enabled])
+        AC_ARG_ENABLE( [debug],
+                AS_HELP_STRING([--enable-debug],
+                [Enable generic debug support (default off)]),
+                [ case "$enableval" in
+                        yes) zfs_ac_debug=yes ;;
+                        no)  zfs_ac_debug=no  ;;
+                        *) AC_MSG_RESULT([Error!])
+                        AC_MSG_ERROR([Bad value "$enableval" for --enable-debug]) ;;
+                esac ]
+        )
+        if test "$zfs_ac_debug" = yes; then
+        	AC_MSG_RESULT([yes])
+                KERNELCPPFLAGS="${KERNELCPPFLAGS} -DDEBUG "
+		HOSTCFLAGS="${HOSTCFLAGS} -DDEBUG "
+        else
+        	AC_MSG_RESULT([no])
+                AC_DEFINE([NDEBUG], [1],
+                [Define to 1 to disable debug tracing])
+                KERNELCPPFLAGS="${KERNELCPPFLAGS} -DNDEBUG "
+		HOSTCFLAGS="${HOSTCFLAGS} -DNDEBUG "
+        fi
+])
+
+AC_DEFUN([ZFS_AC_SCRIPT_CONFIG], [
+	SCRIPT_CONFIG=.script-config
+	rm -f ${SCRIPT_CONFIG}
+	echo "KERNELSRC=${LINUX}"         >>${SCRIPT_CONFIG}
+	echo "KERNELBUILD=${LINUX_OBJ}"   >>${SCRIPT_CONFIG}
+	echo "KERNELSRCVER=$kernsrcver"   >>${SCRIPT_CONFIG}
+	echo                              >>${SCRIPT_CONFIG}
+	echo "SPLSRC=$splsrc"             >>${SCRIPT_CONFIG}
+	echo "SPLBUILD=$splbuild"         >>${SCRIPT_CONFIG}
+	echo "SPLSRCVER=$splsrcver"       >>${SCRIPT_CONFIG}
+	echo "SPLSYMVERS=$splsymvers"     >>${SCRIPT_CONFIG}
+	echo                              >>${SCRIPT_CONFIG}
+	echo "ZFSSRC=${TOPDIR}/src"       >>${SCRIPT_CONFIG}
+	echo "ZFSBUILD=${ZFSDIR}"         >>${SCRIPT_CONFIG}
+	echo                              >>${SCRIPT_CONFIG}
+	echo "TOPDIR=${TOPDIR}            >>${SCRIPT_CONFIG}
+	echo "LIBDIR=${LIBDIR}            >>${SCRIPT_CONFIG}
+	echo "CMDDIR=${CMDDIR}            >>${SCRIPT_CONFIG}
+])
+
+dnl #
+dnl # ZFS_LINUX_CONFTEST
+dnl #
+AC_DEFUN([ZFS_LINUX_CONFTEST], [
+cat >conftest.c <<_ACEOF
+$1
+_ACEOF
+])
+
+dnl #
+dnl # ZFS_LANG_PROGRAM(C)([PROLOGUE], [BODY])
+dnl #
+m4_define([ZFS_LANG_PROGRAM], [
+$1
+int
+main (void)
+{
+dnl Do *not* indent the following line: there may be CPP directives.
+dnl Don't move the `;' right after for the same reason.
+$2
+  ;
+  return 0;
+}
+])
+
+dnl #
+dnl # ZFS_LINUX_COMPILE_IFELSE / like AC_COMPILE_IFELSE
+dnl #
+AC_DEFUN([ZFS_LINUX_COMPILE_IFELSE], [
+m4_ifvaln([$1], [ZFS_LINUX_CONFTEST([$1])])dnl
+rm -f build/conftest.o build/conftest.mod.c build/conftest.ko build/Makefile
+echo "obj-m := conftest.o" >build/Makefile
+dnl AS_IF([AC_TRY_COMMAND(cp conftest.c build && make [$2] CC="$CC" -f $PWD/build/Makefile LINUXINCLUDE="-Iinclude -include include/linux/autoconf.h" -o tmp_include_depends -o scripts -o include/config/MARKER -C $LINUX_OBJ EXTRA_CFLAGS="-Werror-implicit-function-declaration $EXTRA_KCFLAGS" $ARCH_UM SUBDIRS=$PWD/build) >/dev/null && AC_TRY_COMMAND([$3])],
+AS_IF([AC_TRY_COMMAND(cp conftest.c build && make [$2] CC="$CC" LINUXINCLUDE="-Iinclude -include include/linux/autoconf.h" -o tmp_include_depends -o scripts -o include/config/MARKER -C $LINUX_OBJ EXTRA_CFLAGS="-Werror-implicit-function-declaration $EXTRA_KCFLAGS" $ARCH_UM M=$PWD/build) >/dev/null && AC_TRY_COMMAND([$3])],
+        [$4],
+        [_AC_MSG_LOG_CONFTEST
+m4_ifvaln([$5],[$5])dnl])dnl
+rm -f build/conftest.o build/conftest.mod.c build/conftest.mod.o build/conftest.ko m4_ifval([$1], [build/conftest.c conftest.c])[]dnl
+])
+
+dnl #
+dnl # ZFS_LINUX_TRY_COMPILE like AC_TRY_COMPILE
+dnl #
+AC_DEFUN([ZFS_LINUX_TRY_COMPILE],
+	[ZFS_LINUX_COMPILE_IFELSE(
+        [AC_LANG_SOURCE([ZFS_LANG_PROGRAM([[$1]], [[$2]])])],
+        [modules],
+        [test -s build/conftest.o],
+        [$3], [$4])
+])
+
+dnl #
+dnl # ZFS_LINUX_CONFIG
+dnl #
+AC_DEFUN([ZFS_LINUX_CONFIG],
+	[AC_MSG_CHECKING([whether Linux was built with CONFIG_$1])
+	ZFS_LINUX_TRY_COMPILE([
+		#ifndef AUTOCONF_INCLUDED
+		#include <linux/config.h>
+		#endif
+	],[
+		#ifndef CONFIG_$1
+		#error CONFIG_$1 not #defined
+		#endif
+	],[
+		AC_MSG_RESULT([yes])
+		$2
+	],[
+		AC_MSG_RESULT([no])
+		$3
+	])
+])
+
+dnl #
+dnl # ZFS_CHECK_SYMBOL_EXPORT
+dnl # check symbol exported or not
+dnl #
+AC_DEFUN([ZFS_CHECK_SYMBOL_EXPORT],
+	[AC_MSG_CHECKING([whether symbol $1 is exported])
+	grep -q -E '[[[:space:]]]$1[[[:space:]]]' $LINUX/Module.symvers 2>/dev/null
+	rc=$?
+	if test $rc -ne 0; then
+		export=0
+		for file in $2; do
+			grep -q -E "EXPORT_SYMBOL.*($1)" "$LINUX/$file" 2>/dev/null
+			rc=$?
+		        if test $rc -eq 0; then
+		                export=1
+		                break;
+		        fi
+		done
+		if test $export -eq 0; then
+			AC_MSG_RESULT([no])
+			$4
+		else
+			AC_MSG_RESULT([yes])
+			$3
+		fi
+	else
+		AC_MSG_RESULT([yes])
+		$3
+	fi
+])
+
+dnl #
+dnl # 2.6.x API change
+dnl # bio_end_io_t uses 2 args (size was dropped from prototype)
+dnl #
+AC_DEFUN([ZFS_AC_2ARGS_BIO_END_IO_T],
+	[AC_MSG_CHECKING([whether bio_end_io_t wants 2 args])
+	tmp_flags="$EXTRA_KCFLAGS"
+	EXTRA_KCFLAGS="-Werror"
+	ZFS_LINUX_TRY_COMPILE([
+	        #include <linux/bio.h>
+	],[
+		void (*wanted_end_io)(struct bio *, int) = NULL;
+		bio_end_io_t *local_end_io;
+
+		local_end_io = wanted_end_io;
+	],[
+	        AC_MSG_RESULT(yes)
+	        AC_DEFINE(HAVE_2ARGS_BIO_END_IO_T, 1,
+	                  [bio_end_io_t wants 2 args])
+	],[
+	        AC_MSG_RESULT(no)
+	])
+	EXTRA_KCFLAGS="$tmp_flags"
+])
--- a/autogen.sh
+++ b/autogen.sh
@ -0,0 +1,10 @@
+#!/bin/sh
+
+find . -type d -name .deps | xargs rm -rf
+rm -rf config.guess config.sub ltmain.sh
+libtoolize --automake
+aclocal -I autoconf 2>/dev/null &&
+autoheader &&
+automake --add-missing --include-deps # 2>/dev/null &&
+autoconf
+
--- a/build/Makefile
+++ b/build/Makefile
@ -0,0 +1 @@
+obj-m := conftest.o
--- a/configs/Makefile.am
+++ b/configs/Makefile.am
@ -0,0 +1 @@
+EXTRA_DIST=kernel user lustre
--- a/configs/kernel
+++ b/configs/kernel
@ -0,0 +1,11 @@
+# Default ZFS kernel mode configuration 
+
+UNAME=`uname -r | cut -d- -f1`
+
+CONFIG=user
+
+NAME=zfs
+BRANCH=`awk '/[Bb]ranch:/ {print $$2}' META`
+VERSION=`awk '/[Vv]ersion:/ {print $$2}' META`
+RELEASE=`awk '/[Rr]elease:/ {print $$2}' META`
+BUILDDIR=$NAME+$CONFIG
--- a/configs/lustre
+++ b/configs/lustre
@ -0,0 +1,11 @@
+# Default ZFS lustre mode configuration
+
+UNAME=`uname -r | cut -d- -f1`
+
+CONFIG=lustre
+
+NAME=zfs
+BRANCH=`awk '/[Bb]ranch:/ {print $$2}' META`
+VERSION=`awk '/[Vv]ersion:/ {print $$2}' META`
+RELEASE=`awk '/[Rr]elease:/ {print $$2}' META`
+BUILDDIR=$NAME+$CONFIG
--- a/configs/user
+++ b/configs/user
@ -0,0 +1,9 @@
+# Default ZFS user mode configuration
+
+CONFIG=user
+
+NAME=zfs
+BRANCH=`awk '/[Bb]ranch:/ {print $$2}' META`
+VERSION=`awk '/[Vv]ersion:/ {print $$2}' META`
+RELEASE=`awk '/[Rr]elease:/ {print $$2}' META`
+BUILDDIR=$NAME+$CONFIG
--- a/configure.ac
+++ b/configure.ac
@ -0,0 +1,223 @@
+#
+# This file is part of the ZFS Linux port.
+#
+# Copyright (c) 2008 Lawrence Livermore National Security, LLC.
+# Produced at Lawrence Livermore National Laboratory
+# Written by:
+#         Brian Behlendorf <behlendorf1@llnl.gov>,
+#         Herb Wartens <wartens2@llnl.gov>,
+#         Jim Garlick <garlick@llnl.gov>
+# LLNL-CODE-403049
+#
+# CDDL HEADER START
+#
+# The contents of this file are subject to the terms of the
+# Common Development and Distribution License, Version 1.0 only
+# (the "License").  You may not use this file except in compliance
+# with the License.
+#
+# You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
+# or http://www.opensolaris.org/os/licensing.
+# See the License for the specific language governing permissions
+# and limitations under the License.
+#
+# When distributing Covered Code, include this CDDL HEADER in each
+# file and include the License file at usr/src/OPENSOLARIS.LICENSE.
+# If applicable, add the following below this CDDL HEADER, with the
+# fields enclosed by brackets "[]" replaced with your own identifying
+# information: Portions Copyright [yyyy] [name of copyright owner]
+#
+# CDDL HEADER END
+#
+
+AC_INIT
+AC_LANG(C)
+
+AC_CANONICAL_SYSTEM
+AM_INIT_AUTOMAKE(zfs, 0.4.0)
+AC_CONFIG_HEADERS([zfs_config.h])
+
+AC_PROG_INSTALL
+AC_PROG_CC
+AC_PROG_LIBTOOL
+
+zfsconfig=kernel
+kernelsrc=
+kernelbuild=
+splsrc=
+splbuild=
+
+ZFS_AC_CONFIG
+ZFS_AC_KERNEL
+ZFS_AC_SPL
+ZFS_AC_SCRIPT_CONFIG
+ZFS_AC_LICENSE
+ZFS_AC_DEBUG
+ZFS_AC_2ARGS_BIO_END_IO_T
+
+AC_SUBST(UNAME)
+AC_SUBST(CONFIG)
+AC_SUBST(NAME)
+AC_SUBST(SVNURL)
+AC_SUBST(BRANCH)
+AC_SUBST(VERSION)
+AC_SUBST(RELEASE)
+AC_SUBST(BRANCHURL)
+AC_SUBST(TAGURL)
+AC_SUBST(BUILDURL)
+AC_SUBST(BUILDDIR)
+
+# Check for needed userspace bits
+AC_CHECK_HEADERS(sys/types.h sys/byteorder.h sys/isa_defs.h \
+                 sys/systeminfo.h sys/u8_textprep.h libdiskmgt.h)
+
+AC_CHECK_FUNCS(strlcat strlcpy strnlen issetugid setmntent getexecname)
+
+AC_CHECK_LIB([diskmgt], [libdiskmgt_error],
+	[AC_DEFINE([HAVE_LIBDISKMGT], 1,
+        [Define to 1 if 'libdiskmgt' library available])])
+
+AC_CHECK_LIB([efi], [efi_alloc_and_init],
+	[AC_DEFINE([HAVE_LIBEFI], 1,
+	[Define to 1 if 'libefi' library available])])
+
+AC_CHECK_LIB([share], [sa_init],
+	[AC_DEFINE([HAVE_LIBSHARE], 1,
+	[Define to 1 if 'libshare' library available])])
+
+AC_EGREP_HEADER(ioctl, unistd.h,
+	[AC_DEFINE([HAVE_IOCTL_IN_UNISTD_H], 1,
+        [Define to 1 if ioctl() is defined in <unistd.h> header file])])
+
+AC_EGREP_HEADER(ioctl, sys/ioctl.h,
+	[AC_DEFINE([HAVE_IOCTL_IN_SYS_IOCTL_H], 1,
+        [Define to 1 if ioctl() is defined in <sys/ioctl.h> header file])])
+
+AC_EGREP_HEADER(ioctl, stropts.h,
+	[AC_DEFINE([HAVE_IOCTL_IN_STROPTS_H], 1,
+        [Define to 1 if ioctl() is defined in <stropts.h> header file])])
+
+AC_EGREP_HEADER(strcmp, strings.h, 
+	[AC_DEFINE([HAVE_STRCMP_IN_STRINGS_H], 1,
+        [Define to 1 if strcmpl() is defined in <strings.h> header file])])
+
+AC_EGREP_HEADER(sysinfo, sys/systeminfo.h, 
+	[AC_DEFINE([HAVE_SYSINFO_IN_SYS_SYSTEMINFO_H], 1,
+        [Define to 1 if sysinfo() is defined in <sys/systeminfo.h> header file])])
+
+#AC_DEFINE([HAVE_ZVOL], 1, ["Define to 1 to include ZVOL support"])
+#AC_DEFINE([HAVE_ZPL], 1, ["Define to 1 to include ZPL support"])
+#AC_DEFINE([WANT_FAKE_IOCTL], 1, ["Define to 1 to use fake ioctl() support"])
+#AC_DEFINE([HAVE_DM_INUSE_SWAP], 1, ["None"])
+#AC_DEFINE([HAVE_UNICODE], 1, ["None"])
+#AC_DEFINE([HAVE_INTTYPES], 1, [Define to 1 if unint16 defined in <sys/types.h> header file])
+
+# Add "V=1" to KERNELMAKE_PARAMS to enable verbose module build
+KERNELMAKE_PARAMS=
+KERNELCPPFLAGS="$KERNELCPPFLAGS -DHAVE_SPL -D_KERNEL -I$splsrc -I$splsrc/include -I$TOPDIR"
+
+# Minimally required for pread() functionality an other GNU goodness
+HOSTCFLAGS="$HOSTCFLAGS -ggdb -O2 -std=c99 -D_GNU_SOURCE -D__EXTENSIONS__ "
+# Quiet warnings not covered by the gcc-* patches
+HOSTCFLAGS="$HOSTCFLAGS -Wno-switch -Wno-unused -Wno-missing-braces -Wno-parentheses "
+HOSTCFLAGS="$HOSTCFLAGS -Wno-uninitialized -fno-strict-aliasing "
+# Expected defines not covered by zfs_config.h
+HOSTCFLAGS="$HOSTCFLAGS -DHAVE_SPL -D_POSIX_PTHREAD_SEMANTICS "
+HOSTCFLAGS="$HOSTCFLAGS -D_FILE_OFFSET_BITS=64 -D_LARGEFILE64_SOURCE -D_REENTRANT "
+HOSTCFLAGS="$HOSTCFLAGS -DTEXT_DOMAIN=\\\"zfs-linux-kernel\\\" "
+# Expected default include paths additional paths added by Makefiles
+HOSTCFLAGS="$HOSTCFLAGS -I$TOPDIR "
+
+if test "$kernelbuild" != "$kernelsrc"; then
+	KERNELMAKE_PARAMS="$KERNELMAKE_PARAMS O=$kernelbuild"
+fi
+
+AC_SUBST(KERNELMAKE_PARAMS)
+AC_SUBST(KERNELCPPFLAGS)
+AC_SUBST(HOSTCFLAGS)
+
+AC_CONFIG_FILES([ Makefile
+                  autoconf/Makefile
+                  configs/Makefile
+                  doc/Makefile
+                  scripts/Makefile
+                  zfs/Makefile
+                  zfs/lib/libudmu/include/Makefile
+                  zfs/lib/libudmu/Makefile
+                  zfs/lib/Makefile
+                  zfs/lib/libnvpair/include/sys/Makefile
+                  zfs/lib/libnvpair/include/Makefile
+                  zfs/lib/libnvpair/Makefile
+                  zfs/lib/libsolcompat/sparc64/Makefile
+                  zfs/lib/libsolcompat/Makefile
+                  zfs/lib/libsolcompat/include/tsol/Makefile
+                  zfs/lib/libsolcompat/include/sparc64/sys/Makefile
+                  zfs/lib/libsolcompat/include/sparc64/Makefile
+                  zfs/lib/libsolcompat/include/rpc/Makefile
+                  zfs/lib/libsolcompat/include/i386/sys/Makefile
+                  zfs/lib/libsolcompat/include/i386/Makefile
+                  zfs/lib/libsolcompat/include/ia32/sys/Makefile
+                  zfs/lib/libsolcompat/include/ia32/Makefile
+                  zfs/lib/libsolcompat/include/amd64/sys/Makefile
+                  zfs/lib/libsolcompat/include/amd64/Makefile
+                  zfs/lib/libsolcompat/include/sys/sysevent/Makefile
+                  zfs/lib/libsolcompat/include/sys/fm/Makefile
+                  zfs/lib/libsolcompat/include/sys/Makefile
+                  zfs/lib/libsolcompat/include/Makefile
+                  zfs/lib/libsolcompat/i386/Makefile
+                  zfs/lib/libsolcompat/amd64/Makefile
+                  zfs/lib/libavl/include/sys/Makefile
+                  zfs/lib/libavl/include/Makefile
+                  zfs/lib/libavl/Makefile
+                  zfs/lib/libuutil/include/Makefile
+                  zfs/lib/libuutil/Makefile
+                  zfs/lib/libzfs/include/Makefile
+                  zfs/lib/libzfs/Makefile
+                  zfs/lib/libumem/include/Makefile
+                  zfs/lib/libumem/Makefile
+                  zfs/lib/libumem/sys/Makefile
+                  zfs/lib/libzcommon/include/Makefile
+                  zfs/lib/libzcommon/include/sys/fm/fs/Makefile
+                  zfs/lib/libzcommon/include/sys/fm/Makefile
+                  zfs/lib/libzcommon/include/sys/Makefile
+                  zfs/lib/libzcommon/include/sys/fs/Makefile
+                  zfs/lib/libzcommon/Makefile
+                  zfs/lib/libzpool/Makefile
+                  zfs/lib/libport/include/sys/Makefile
+                  zfs/lib/libport/include/Makefile
+                  zfs/lib/libport/Makefile
+                  zfs/lib/libdmu-ctl/include/sys/Makefile
+                  zfs/lib/libdmu-ctl/include/Makefile
+                  zfs/lib/libdmu-ctl/Makefile
+                  zfs/zcmd/ztest/Makefile
+                  zfs/zcmd/Makefile
+                  zfs/zcmd/zfs/Makefile
+                  zfs/zcmd/zdb/Makefile
+                  zfs/zcmd/zinject/Makefile
+                  zfs/zcmd/zdump/Makefile
+                  zfs/zcmd/zpool/Makefile
+               ])
+AC_OUTPUT
+
+# HACK: I really, really hate this...  but to ensure the kernel build 
+# system compiles C files shared between a library and a kernel module, 
+# we need to ensure each file has a unique make target.  To do that 
+# I'm creating symlinks for each shared file at configure time.  It 
+# may be possible something better can be done in the Makefile but it
+# will take some serious investigation and I don't have the time now.
+
+echo
+echo "Creating symlinks for additional make targets"
+ln -s $LIBDIR/libport/u8_textprep.c      $LIBDIR/libport/ku8_textprep.c
+ln -s $LIBDIR/libavl/avl.c               $LIBDIR/libavl/kavl.c
+ln -s $LIBDIR/libavl/avl.c               $LIBDIR/libavl/uavl.c
+ln -s $LIBDIR/libnvpair/nvpair.c         $LIBDIR/libnvpair/knvpair.c
+ln -s $LIBDIR/libnvpair/nvpair.c         $LIBDIR/libnvpair/unvpair.c
+ln -s $LIBDIR/libzcommon/zfs_deleg.c     $LIBDIR/libzcommon/kzfs_deleg.c
+ln -s $LIBDIR/libzcommon/zfs_prop.c      $LIBDIR/libzcommon/kzfs_prop.c
+ln -s $LIBDIR/libzcommon/zprop_common.c  $LIBDIR/libzcommon/kzprop_common.c
+ln -s $LIBDIR/libzcommon/compress.c      $LIBDIR/libzcommon/kcompress.c
+ln -s $LIBDIR/libzcommon/list.c          $LIBDIR/libzcommon/klist.c
+ln -s $LIBDIR/libzcommon/zfs_namecheck.c $LIBDIR/libzcommon/kzfs_namecheck.c
+ln -s $LIBDIR/libzcommon/zfs_comutil.c   $LIBDIR/libzcommon/kzfs_comutil.c
+ln -s $LIBDIR/libzcommon/zpool_prop.c    $LIBDIR/libzcommon/kzpool_prop.c
--- a/doc/LEGAL
+++ b/doc/LEGAL
@ -0,0 +1,113 @@
+From: Chris Dunlap <cdunlap@llnl.gov>
+To: tak1@llnl.gov (James Tak)
+Cc: rogers11@llnl.gov (Leah Rogers), garlick@llnl.gov (Jim Garlick),
+        mgary@llnl.gov (Mark Gary), kimcupps@llnl.gov (Kim Cupps)
+Date: Mon, 26 Mar 2007 15:37:07 -0700
+Subject: CDDL/GPL licensing issues for ZFS Linux port
+
+James,
+
+We want to port Sun's Zettabyte File System (ZFS) to Linux and
+ultimately redistribute the source code of our work.  We've been
+talking with Leah about this and have a meeting scheduled with you
+for this coming Thursday at 2pm.  I just wanted to give you a summary
+before the meeting of what we're proposing.
+
+ZFS is part of OpenSolaris which is licensed under the Common
+Development and Distribution License (CDDL):
+
+  http://www.opensolaris.org/os/licensing/cddllicense.txt
+
+The Linux kernel is licensed under the GNU General Public License (GPL)
+(specifically, under version 2 of the license only):
+
+  http://www.fsf.org/licensing/licenses/gpl.html
+
+While these are both Open-Source licenses, the Free Software Foundation
+(FSF) states they are incompatible with one another:
+
+  http://www.fsf.org/licensing/licenses/index_html
+
+  "[CDDL] is a free software license which is not a strong copyleft;
+  it has some complex restrictions that make it incompatible with the
+  GNU GPL.  It requires that all attribution notices be maintained,
+  while the GPL only requires certain types of notices.  Also, it
+  terminates in retaliation for certain aggressive uses of patents.
+  So, a module covered by the GPL and a module covered by the CDDL
+  cannot legally be linked together."
+
+As an aside, Sun is reportedly considering releasing OpenSolaris under
+GPL3 (i.e., the upcoming version 3 of the GNU General Public License):
+
+  http://blogs.sun.com/jonathan/entry/hp_and_sun_partnering_around
+
+  http://arstechnica.com/news.ars/post/20060130-6074.html
+
+  http://news.com.com/Sun+considers+GPL+3+license+for+Solaris/2100-1016_3-6032893.html
+
+Since the GPL3 has not been finalized, it is unclear whether
+incompatibilities will exist between GPL2 and GPL3.
+
+Linus Torvalds (the original creator of Linux) describes his views
+on the licensing of Linux kernel modules in the following email thread:
+
+  http://linuxmafia.com/faq/Kernel/proprietary-kernel-modules.html
+
+Most of this thread is in regards to proprietary closed-source
+binary-only modules for Linux.  Linus generally considers modules
+written for Linux using the kernel infrastructures to be derived
+works of Linux, even if they don't copy any existing Linux code.
+However, he specifically singles out drivers and filesystems ported
+from other operating systems as not being derived works:
+
+  "It would be rather preposterous to call the Andrew FileSystem a
+  'derived work' of Linux, for example, so I think it's perfectly
+  OK to have a AFS module, for example."
+
+  "The original binary-only modules were for things that were
+  pre-existing works of code, i.e., drivers and filesystems ported
+  from other operating systems, which thus could clearly be argued
+  to not be derived works..."
+
+Based on this, it seems our port of Sun's ZFS filesystem to Linux
+would not be considered a derived work of Linux, and therefore not
+covered by the GPL.  The issue of the CDDL/GPL license incompatibility
+becomes moot.  As such, we should be able to redistribute our changes
+to ZFS in source-code form licensed under the CDDL since this will
+be a derived work of the original ZFS code.  There seems to be some
+dissent as to whether a binary module could be redistributed as well,
+but that issue does not concern us.  In this instance, we are only
+interested in redistribution of our work in source-code form.
+
+-Chris
+
+To: Chris Dunlap <cdunlap@llnl.gov>
+From: James Tak <tak1@llnl.gov>
+Subject: Re: CDDL/GPL licensing issues for ZFS Linux port
+Cc: rogers11@llnl.gov (Leah Rogers), garlick@llnl.gov (Jim Garlick),
+        mgary@llnl.gov (Mark Gary), kimcupps@llnl.gov (Kim Cupps)
+Date: Thu, 29 Mar 2007 14:53:01 -0700
+
+Hi Chris,
+As per our discussion today, the ZFS port you are proposing releasing under
+the CDDL license should be o.k. since it is a derivative work of the
+original ZFS module (under CDDL) and is therefore also subject to CDDL
+under the distribution terms of that license.  While the issue of linking
+has been greatly debated in the OS community, I think it is fair to say in
+this instance the ZFS port is not a derivative work of Linux and thus not
+subject to the GPL.  Furthermore, it shouldn't be a problem especially
+since even Linus Torvald has expressed that modules such as yours are not
+derived works of Linux.
+
+Let me know if you have any further questions at x27274.  Thanks.
+
+Regards,
+James
+
+James S. Tak
+Assistant Laboratory Counsel for Intellectual Property
+Office of Laboratory Counsel
+Lawrence Livermore National Laboratory
+phone: (925) 422-7274
+fax: (925) 423-2231
+tak1@llnl.gov
--- a/doc/Makefile.am
+++ b/doc/Makefile.am
@ -0,0 +1 @@
+EXTRA_DIST = LEGAL
--- a/doc/analysis-spl-0.3.1/zfs-report.doc
+++ b/doc/analysis-spl-0.3.1/zfs-report.doc
--- a/doc/analysis-spl-0.3.1/zfs-report.odt
+++ b/doc/analysis-spl-0.3.1/zfs-report.odt
--- a/doc/analysis-spl-0.3.1/zfs-report.pdf
+++ b/doc/analysis-spl-0.3.1/zfs-report.pdf
--- a/doc/analysis-spl-0.3.1/zfs-testing.ods
+++ b/doc/analysis-spl-0.3.1/zfs-testing.ods
--- a/doc/analysis-spl-0.3.1/zfs-testing.xls
+++ b/doc/analysis-spl-0.3.1/zfs-testing.xls
--- a/lib/libsolcompat/zu.c
+++ b/lib/libsolcompat/zu.c
@ -0,0 +1,4 @@
+int main(void)
+{
+	return 0;
+}
--- a/lib/libumem/zu.c
+++ b/lib/libumem/zu.c
@ -0,0 +1,4 @@
+int main(void)
+{
+	return 0;
+}
--- a/lib/libuutil/zu.c
+++ b/lib/libuutil/zu.c
@ -0,0 +1,4 @@
+int main(void)
+{
+	return 0;
+}
--- a/lib/libzfs/zu.c
+++ b/lib/libzfs/zu.c
@ -0,0 +1,4 @@
+int main(void)
+{
+	return 0;
+}
--- a/patches/README
+++ b/patches/README
@ -0,0 +1,17 @@
+#
+# Mostly patches for a userspace build.  For now I'm leaving them all
+# out until we go through the code base and sort out the userspace
+# portion of the build system.  We may find we do not want or need
+# many of these patches anymore. -Brian
+#
+zap-cursor-move-to-key.patch    # Add a ZAP API to move a ZAP cursor to a
+given key.
+spa-force-readonly.patch        # Add API to discard all writes
+no-debug-userspace.patch        # Disable debug code on userspace
+no-events.patch                 # Define away spa_event_notify() in userspace
+pthreads.patch                  # Use POSIX threads in userspace.
+port-no-zmod.patch              # Do not use zmod.h in userspace.
+port-pragma-init.patch          # Use constructor attribute on non-Solaris
+platforms.
+lztest-lzdb.patch               # Make lztest call lzdb from PATH.
+zpool-force.patch               # Change -f to -F in zpool command
--- a/patches/lztest-lzdb.patch
+++ b/patches/lztest-lzdb.patch
@ -0,0 +1,40 @@
+Make lztest call lzdb from PATH.
+
+Index: zfs+chaos4/cmd/lztest/ztest.c
+===================================================================
+--- zfs+chaos4.orig/cmd/lztest/ztest.c
+++ zfs+chaos4/cmd/lztest/ztest.c
+@@ -3043,30 +3043,17 @@ ztest_verify_blocks(char *pool)
+ 	char zbuf[1024];
+ 	char *bin;
+ 	char *ztest;
+-	char *isa;
+-	int isalen;
+ 	FILE *fp;
+ 
+-	(void) realpath(getexecname(), zdb);
+-
+-	/* zdb lives in /usr/sbin, while ztest lives in /usr/bin */
+-	bin = strstr(zdb, "/usr/bin/");
+-	ztest = strstr(bin, "/ztest");
+-	isa = bin + 8;
+-	isalen = ztest - isa;
+-	isa = strdup(isa);
+ 	/* LINTED */
+-	(void) sprintf(bin,
+-	    "/usr/sbin%.*s/zdb -bc%s%s -U /tmp/zpool.cache -O %s %s",
+-	    isalen,
+-	    isa,
+	(void) sprintf(zdb,
+	    "lzdb -bc%s%s -U /tmp/zpool.cache -O %s %s",
+ 	    zopt_verbose >= 3 ? "s" : "",
+ 	    zopt_verbose >= 4 ? "v" : "",
+ 	    ztest_random(2) == 0 ? "pre" : "post", pool);
+-	free(isa);
+ 
+ 	if (zopt_verbose >= 5)
+-		(void) printf("Executing %s\n", strstr(zdb, "zdb "));
+		(void) printf("Executing %s\n", strstr(zdb, "lzdb "));
+ 
+ 	fp = popen(zdb, "r");
+ 
--- a/patches/no-debug-userspace.patch
+++ b/patches/no-debug-userspace.patch
@ -0,0 +1,184 @@
+Disable debug code on userspace
+
+Index: zfs+chaos4/lib/libzfscommon/include/sys/arc.h
+===================================================================
+--- zfs+chaos4.orig/lib/libzfscommon/include/sys/arc.h
+++ zfs+chaos4/lib/libzfscommon/include/sys/arc.h
+@@ -82,7 +82,7 @@ int arc_released(arc_buf_t *buf);
+ int arc_has_callback(arc_buf_t *buf);
+ void arc_buf_freeze(arc_buf_t *buf);
+ void arc_buf_thaw(arc_buf_t *buf);
+-#ifdef ZFS_DEBUG
+#if defined(ZFS_DEBUG) || (!defined(_KERNEL) && !defined(NDEBUG))
+ int arc_referenced(arc_buf_t *buf);
+ #endif
+ 
+Index: zfs+chaos4/lib/libzfscommon/include/sys/refcount.h
+===================================================================
+--- zfs+chaos4.orig/lib/libzfscommon/include/sys/refcount.h
+++ zfs+chaos4/lib/libzfscommon/include/sys/refcount.h
+@@ -43,7 +43,7 @@ extern "C" {
+  */
+ #define	FTAG ((char *)__func__)
+ 
+-#if defined(DEBUG) || !defined(_KERNEL)
+#if defined(DEBUG)
+ typedef struct reference {
+ 	list_node_t ref_link;
+ 	void *ref_holder;
+Index: zfs+chaos4/lib/libzfscommon/include/sys/zfs_context_user.h
+===================================================================
+--- zfs+chaos4.orig/lib/libzfscommon/include/sys/zfs_context_user.h
+++ zfs+chaos4/lib/libzfscommon/include/sys/zfs_context_user.h
+@@ -96,6 +96,8 @@ extern "C" {
+ 
+ #ifdef ZFS_DEBUG
+ extern void dprintf_setup(int *argc, char **argv);
+#else
+#define dprintf_setup(ac,av) ((void) 0)
+ #endif /* ZFS_DEBUG */
+ 
+ extern void cmn_err(int, const char *, ...);
+@@ -105,21 +107,26 @@ extern void vpanic(const char *, __va_li
+ 
+ #define	fm_panic	panic
+ 
+#ifndef zp_verify
+ /* This definition is copied from assert.h. */
+ #if defined(__STDC__)
+ #if __STDC_VERSION__ - 0 >= 199901L
+-#define	verify(EX) (void)((EX) || \
+#define	zp_verify(EX) (void)((EX) || \
+ 	(__assert_c99(#EX, __FILE__, __LINE__, __func__), 0))
+ #else
+-#define	verify(EX) (void)((EX) || (__assert(#EX, __FILE__, __LINE__), 0))
+#define	zp_verify(EX) (void)((EX) || (__assert(#EX, __FILE__, __LINE__), 0))
+ #endif /* __STDC_VERSION__ - 0 >= 199901L */
+ #else
+-#define	verify(EX) (void)((EX) || (_assert("EX", __FILE__, __LINE__), 0))
+#define	zp_verify(EX) (void)((EX) || (_assert("EX", __FILE__, __LINE__), 0))
+ #endif	/* __STDC__ */
+#endif
+ 
+-
+-#define	VERIFY	verify
+#ifndef VERIFY
+#define	VERIFY	zp_verify
+#endif
+#ifndef ASSERT
+ #define	ASSERT	assert
+#endif
+ 
+ extern void __assert(const char *, const char *, int);
+ 
+@@ -332,6 +339,7 @@ extern int	taskq_member(taskq_t *, void 
+ typedef struct vnode {
+ 	uint64_t	v_size;
+ 	int		v_fd;
+	mode_t		v_mode;
+ 	char		*v_path;
+ } vnode_t;
+ 
+Index: zfs+chaos4/lib/libzfscommon/include/sys/zfs_debug.h
+===================================================================
+--- zfs+chaos4.orig/lib/libzfscommon/include/sys/zfs_debug.h
+++ zfs+chaos4/lib/libzfscommon/include/sys/zfs_debug.h
+@@ -44,7 +44,7 @@ extern "C" {
+  * ZFS debugging
+  */
+ 
+-#if defined(DEBUG) || !defined(_KERNEL)
+#if defined(DEBUG)
+ #define	ZFS_DEBUG
+ #endif
+ 
+Index: zfs+chaos4/lib/libzpool/arc.c
+===================================================================
+--- zfs+chaos4.orig/lib/libzpool/arc.c
+++ zfs+chaos4/lib/libzpool/arc.c
+@@ -1802,7 +1802,7 @@ arc_reclaim_needed(void)
+ 		return (1);
+ #endif
+ 
+-#else
+#elif defined(ZFS_DEBUG)
+ 	if (spa_get_random(100) == 0)
+ 		return (1);
+ #endif
+@@ -2881,7 +2881,7 @@ arc_has_callback(arc_buf_t *buf)
+ 	return (buf->b_efunc != NULL);
+ }
+ 
+-#ifdef ZFS_DEBUG
+#if defined(ZFS_DEBUG) || (!defined(_KERNEL) && !defined(NDEBUG))
+ int
+ arc_referenced(arc_buf_t *buf)
+ {
+Index: zfs+chaos4/lib/libzpool/kernel.c
+===================================================================
+--- zfs+chaos4.orig/lib/libzpool/kernel.c
+++ zfs+chaos4/lib/libzpool/kernel.c
+@@ -384,6 +384,7 @@ vn_open(char *path, int x1, int flags, i
+ 
+ 	vp->v_fd = fd;
+ 	vp->v_size = st.st_size;
+	vp->v_mode = st.st_mode;
+ 	vp->v_path = spa_strdup(path);
+ 
+ 	return (0);
+@@ -422,10 +423,17 @@ vn_rdwr(int uio, vnode_t *vp, void *addr
+ 		 * To simulate partial disk writes, we split writes into two
+ 		 * system calls so that the process can be killed in between.
+ 		 */
+-		split = (len > 0 ? rand() % len : 0);
+-		iolen = pwrite64(vp->v_fd, addr, split, offset);
+-		iolen += pwrite64(vp->v_fd, (char *)addr + split,
+-		    len - split, offset + split);
+#ifdef ZFS_DEBUG
+		if (!S_ISBLK(vp->v_mode) && !S_ISCHR(vp->v_mode)) {
+			split = (len > 0 ? rand() % len : 0);
+			iolen = pwrite64(vp->v_fd, addr, split, offset);
+			iolen += pwrite64(vp->v_fd, (char *)addr + split,
+			    len - split, offset + split);
+		} else
+			iolen = pwrite64(vp->v_fd, addr, len, offset);
+#else
+		iolen = pwrite64(vp->v_fd, addr, len, offset);
+#endif
+ 	}
+ 
+ 	if (iolen < 0)
+Index: zfs+chaos4/lib/libzpool/refcount.c
+===================================================================
+--- zfs+chaos4.orig/lib/libzpool/refcount.c
+++ zfs+chaos4/lib/libzpool/refcount.c
+@@ -28,7 +28,7 @@
+ #include <sys/zfs_context.h>
+ #include <sys/refcount.h>
+ 
+-#if defined(DEBUG) || !defined(_KERNEL)
+#if defined(DEBUG)
+ 
+ #ifdef _KERNEL
+ int reference_tracking_enable = FALSE; /* runs out of memory too easily */
+Index: zfs+chaos4/lib/libzpool/spa_misc.c
+===================================================================
+--- zfs+chaos4.orig/lib/libzpool/spa_misc.c
+++ zfs+chaos4/lib/libzpool/spa_misc.c
+@@ -178,11 +178,15 @@ kmem_cache_t *spa_buffer_pool;
+ int spa_mode;
+ 
+ #ifdef ZFS_DEBUG
+#ifdef _KERNEL
+ /* Everything except dprintf is on by default in debug builds */
+ int zfs_flags = ~ZFS_DEBUG_DPRINTF;
+ #else
+int zfs_flags = ~0;
+#endif /* _KERNEL */
+#else
+ int zfs_flags = 0;
+-#endif
+#endif /* ZFS_DEBUG */
+ 
+ /*
+  * zfs_recover can be set to nonzero to attempt to recover from
--- a/patches/no-events.patch
+++ b/patches/no-events.patch
@ -0,0 +1,44 @@
+Define away spa_event_notify() in userspace - not necessary and breaks compilation in older Solaris builds.
+
+Index: zfs+chaos4/lib/libzfscommon/include/sys/spa.h
+===================================================================
+--- zfs+chaos4.orig/lib/libzfscommon/include/sys/spa.h
+++ zfs+chaos4/lib/libzfscommon/include/sys/spa.h
+@@ -516,7 +516,11 @@ extern int spa_prop_get(spa_t *spa, nvli
+ extern void spa_prop_clear_bootfs(spa_t *spa, uint64_t obj, dmu_tx_t *tx);
+ 
+ /* asynchronous event notification */
+#ifdef _KERNEL
+ extern void spa_event_notify(spa_t *spa, vdev_t *vdev, const char *name);
+#else
+#define spa_event_notify(s,v,n) ((void) 0)
+#endif
+ 
+ #ifdef ZFS_DEBUG
+ #define	dprintf_bp(bp, fmt, ...) do {				\
+Index: zfs+chaos4/lib/libzpool/spa.c
+===================================================================
+--- zfs+chaos4.orig/lib/libzpool/spa.c
+++ zfs+chaos4/lib/libzpool/spa.c
+@@ -4449,10 +4449,10 @@ spa_has_spare(spa_t *spa, uint64_t guid)
+  * in the userland libzpool, as we don't want consumers to misinterpret ztest
+  * or zdb as real changes.
+  */
+#ifdef _KERNEL
+ void
+ spa_event_notify(spa_t *spa, vdev_t *vd, const char *name)
+ {
+-#ifdef _KERNEL
+ 	sysevent_t		*ev;
+ 	sysevent_attr_list_t	*attr = NULL;
+ 	sysevent_value_t	value;
+@@ -4497,8 +4497,8 @@ done:
+ 	if (attr)
+ 		sysevent_free_attr(attr);
+ 	sysevent_free(ev);
+-#endif
+ }
+#endif
+ 
+ void
+ spa_discard_io(spa_t *spa)
--- a/patches/port-no-zmod.patch
+++ b/patches/port-no-zmod.patch
@ -0,0 +1,112 @@
+Do not use zmod.h in userspace.
+
+Index: zfs+chaos4/lib/libzpool/gzip.c
+===================================================================
+--- zfs+chaos4.orig/lib/libzpool/gzip.c
+++ zfs+chaos4/lib/libzpool/gzip.c
+@@ -28,22 +28,35 @@
+ 
+ #include <sys/debug.h>
+ #include <sys/types.h>
+-#include <sys/zmod.h>
+ 
+ #ifdef _KERNEL
+
+ #include <sys/systm.h>
+-#else
+#include <sys/zmod.h>
+
+typedef size_t zlen_t;
+#define compress_func   z_compress_level
+#define uncompress_func z_uncompress
+
+#else /* _KERNEL */
+
+ #include <strings.h>
+#include <zlib.h>
+
+typedef uLongf zlen_t;
+#define compress_func   compress2
+#define uncompress_func uncompress
+
+ #endif
+ 
+ size_t
+ gzip_compress(void *s_start, void *d_start, size_t s_len, size_t d_len, int n)
+ {
+-	size_t dstlen = d_len;
+	zlen_t dstlen = d_len;
+ 
+ 	ASSERT(d_len <= s_len);
+ 
+-	if (z_compress_level(d_start, &dstlen, s_start, s_len, n) != Z_OK) {
+	if (compress_func(d_start, &dstlen, s_start, s_len, n) != Z_OK) {
+ 		if (d_len != s_len)
+ 			return (s_len);
+ 
+@@ -51,18 +64,18 @@ gzip_compress(void *s_start, void *d_sta
+ 		return (s_len);
+ 	}
+ 
+-	return (dstlen);
+	return ((size_t) dstlen);
+ }
+ 
+ /*ARGSUSED*/
+ int
+ gzip_decompress(void *s_start, void *d_start, size_t s_len, size_t d_len, int n)
+ {
+-	size_t dstlen = d_len;
+	zlen_t dstlen = d_len;
+ 
+ 	ASSERT(d_len >= s_len);
+ 
+-	if (z_uncompress(d_start, &dstlen, s_start, s_len) != Z_OK)
+	if (uncompress_func(d_start, &dstlen, s_start, s_len) != Z_OK)
+ 		return (-1);
+ 
+ 	return (0);
+Index: zfs+chaos4/lib/libzpool/kernel.c
+===================================================================
+--- zfs+chaos4.orig/lib/libzpool/kernel.c
+++ zfs+chaos4/lib/libzpool/kernel.c
+@@ -36,7 +36,6 @@
+ #include <sys/stat.h>
+ #include <sys/processor.h>
+ #include <sys/zfs_context.h>
+-#include <sys/zmod.h>
+ #include <sys/utsname.h>
+ #include <sys/time.h>
+ 
+@@ -876,31 +875,6 @@ kernel_fini(void)
+ 	urandom_fd = -1;
+ }
+ 
+-int
+-z_uncompress(void *dst, size_t *dstlen, const void *src, size_t srclen)
+-{
+-	int ret;
+-	uLongf len = *dstlen;
+-
+-	if ((ret = uncompress(dst, &len, src, srclen)) == Z_OK)
+-		*dstlen = (size_t)len;
+-
+-	return (ret);
+-}
+-
+-int
+-z_compress_level(void *dst, size_t *dstlen, const void *src, size_t srclen,
+-    int level)
+-{
+-	int ret;
+-	uLongf len = *dstlen;
+-
+-	if ((ret = compress2(dst, &len, src, srclen, level)) == Z_OK)
+-		*dstlen = (size_t)len;
+-
+-	return (ret);
+-}
+-
+ /*ARGSUSED*/
+ size_t u8_textprep_str(char *i, size_t *il, char *o, size_t *ol, int nf,
+     size_t vers, int *err)
--- a/patches/port-pragma-init.patch
+++ b/patches/port-pragma-init.patch
@ -0,0 +1,52 @@
+Use constructor attribute on non-Solaris platforms.
+
+Index: zfs+chaos4/lib/libuutil/uu_misc.c
+===================================================================
+--- zfs+chaos4.orig/lib/libuutil/uu_misc.c
+++ zfs+chaos4/lib/libuutil/uu_misc.c
+@@ -251,7 +251,13 @@ uu_release_child(void)
+ 	uu_release();
+ }
+ 
+#ifdef __GNUC__
+static void
+uu_init(void) __attribute__((constructor));
+#else
+ #pragma init(uu_init)
+#endif
+
+ static void
+ uu_init(void)
+ {
+Index: zfs+chaos4/lib/libzfs/libzfs_mount.c
+===================================================================
+--- zfs+chaos4.orig/lib/libzfs/libzfs_mount.c
+++ zfs+chaos4/lib/libzfs/libzfs_mount.c
+@@ -128,7 +128,13 @@ zfs_share_proto_t share_all_proto[] = {
+ 	PROTO_END
+ };
+ 
+#ifdef __GNUC__
+static void
+zfs_iscsi_init(void) __attribute__((constructor));
+#else
+ #pragma init(zfs_iscsi_init)
+#endif
+
+ static void
+ zfs_iscsi_init(void)
+ {
+@@ -548,8 +554,12 @@ static void (*_sa_update_sharetab_ts)(sa
+  * values to be used later. This is triggered by the runtime loader.
+  * Make sure the correct ISA version is loaded.
+  */
+-
+#ifdef __GNUC__
+static void
+_zfs_init_libshare(void) __attribute__((constructor));
+#else
+ #pragma init(_zfs_init_libshare)
+#endif
+ static void
+ _zfs_init_libshare(void)
+ {
--- a/patches/pthreads.patch
+++ b/patches/pthreads.patch
@ -0,0 +1,924 @@
+Use POSIX threads in userspace.
+
+Index: zfs+chaos4/cmd/lztest/ztest.c
+===================================================================
+--- zfs+chaos4.orig/cmd/lztest/ztest.c
+++ zfs+chaos4/cmd/lztest/ztest.c
+@@ -141,7 +141,7 @@ typedef struct ztest_args {
+ 	spa_t		*za_spa;
+ 	objset_t	*za_os;
+ 	zilog_t		*za_zilog;
+-	thread_t	za_thread;
+	pthread_t	za_thread;
+ 	uint64_t	za_instance;
+ 	uint64_t	za_random;
+ 	uint64_t	za_diroff;
+@@ -224,17 +224,17 @@ ztest_info_t ztest_info[] = {
+  * Stuff we need to share writably between parent and child.
+  */
+ typedef struct ztest_shared {
+-	mutex_t		zs_vdev_lock;
+-	rwlock_t	zs_name_lock;
+-	uint64_t	zs_vdev_primaries;
+-	uint64_t	zs_enospc_count;
+-	hrtime_t	zs_start_time;
+-	hrtime_t	zs_stop_time;
+-	uint64_t	zs_alloc;
+-	uint64_t	zs_space;
+-	ztest_info_t	zs_info[ZTEST_FUNCS];
+-	mutex_t		zs_sync_lock[ZTEST_SYNC_LOCKS];
+-	uint64_t	zs_seq[ZTEST_SYNC_LOCKS];
+	pthread_mutex_t		zs_vdev_lock;
+	pthread_rwlock_t	zs_name_lock;
+	uint64_t		zs_vdev_primaries;
+	uint64_t		zs_enospc_count;
+	hrtime_t		zs_start_time;
+	hrtime_t		zs_stop_time;
+	uint64_t		zs_alloc;
+	uint64_t		zs_space;
+	ztest_info_t		zs_info[ZTEST_FUNCS];
+	pthread_mutex_t		zs_sync_lock[ZTEST_SYNC_LOCKS];
+	uint64_t		zs_seq[ZTEST_SYNC_LOCKS];
+ } ztest_shared_t;
+ 
+ static char ztest_dev_template[] = "%s/%s.%llua";
+@@ -818,7 +818,7 @@ ztest_spa_create_destroy(ztest_args_t *z
+ 	 * Attempt to create an existing pool.  It shouldn't matter
+ 	 * what's in the nvroot; we should fail with EEXIST.
+ 	 */
+-	(void) rw_rdlock(&ztest_shared->zs_name_lock);
+	(void) pthread_rwlock_rdlock(&ztest_shared->zs_name_lock);
+ 	nvroot = make_vdev_root(0, 0, 0, 0, 1);
+ 	error = spa_create(za->za_pool, nvroot, NULL, NULL);
+ 	nvlist_free(nvroot);
+@@ -834,7 +834,7 @@ ztest_spa_create_destroy(ztest_args_t *z
+ 		fatal(0, "spa_destroy() = %d", error);
+ 
+ 	spa_close(spa, FTAG);
+-	(void) rw_unlock(&ztest_shared->zs_name_lock);
+	(void) pthread_rwlock_unlock(&ztest_shared->zs_name_lock);
+ }
+ 
+ /*
+@@ -851,7 +851,7 @@ ztest_vdev_add_remove(ztest_args_t *za)
+ 	if (zopt_verbose >= 6)
+ 		(void) printf("adding vdev\n");
+ 
+-	(void) mutex_lock(&ztest_shared->zs_vdev_lock);
+	(void) pthread_mutex_lock(&ztest_shared->zs_vdev_lock);
+ 
+ 	spa_config_enter(spa, RW_READER, FTAG);
+ 
+@@ -869,7 +869,7 @@ ztest_vdev_add_remove(ztest_args_t *za)
+ 	error = spa_vdev_add(spa, nvroot);
+ 	nvlist_free(nvroot);
+ 
+-	(void) mutex_unlock(&ztest_shared->zs_vdev_lock);
+	(void) pthread_mutex_unlock(&ztest_shared->zs_vdev_lock);
+ 
+ 	if (error == ENOSPC)
+ 		ztest_record_enospc("spa_vdev_add");
+@@ -927,7 +927,7 @@ ztest_vdev_attach_detach(ztest_args_t *z
+ 	int error, expected_error;
+ 	int fd;
+ 
+-	(void) mutex_lock(&ztest_shared->zs_vdev_lock);
+	(void) pthread_mutex_lock(&ztest_shared->zs_vdev_lock);
+ 
+ 	spa_config_enter(spa, RW_READER, FTAG);
+ 
+@@ -1054,7 +1054,7 @@ ztest_vdev_attach_detach(ztest_args_t *z
+ 		    oldpath, newpath, replacing, error, expected_error);
+ 	}
+ 
+-	(void) mutex_unlock(&ztest_shared->zs_vdev_lock);
+	(void) pthread_mutex_unlock(&ztest_shared->zs_vdev_lock);
+ }
+ 
+ /*
+@@ -1071,7 +1071,7 @@ ztest_vdev_LUN_growth(ztest_args_t *za)
+ 	size_t fsize;
+ 	int fd;
+ 
+-	(void) mutex_lock(&ztest_shared->zs_vdev_lock);
+	(void) pthread_mutex_lock(&ztest_shared->zs_vdev_lock);
+ 
+ 	/*
+ 	 * Pick a random leaf vdev.
+@@ -1102,7 +1102,7 @@ ztest_vdev_LUN_growth(ztest_args_t *za)
+ 		(void) close(fd);
+ 	}
+ 
+-	(void) mutex_unlock(&ztest_shared->zs_vdev_lock);
+	(void) pthread_mutex_unlock(&ztest_shared->zs_vdev_lock);
+ }
+ 
+ /* ARGSUSED */
+@@ -1198,7 +1198,7 @@ ztest_dmu_objset_create_destroy(ztest_ar
+ 	uint64_t objects;
+ 	ztest_replay_t zr;
+ 
+-	(void) rw_rdlock(&ztest_shared->zs_name_lock);
+	(void) pthread_rwlock_rdlock(&ztest_shared->zs_name_lock);
+ 	(void) snprintf(name, 100, "%s/%s_temp_%llu", za->za_pool, za->za_pool,
+ 	    (u_longlong_t)za->za_instance);
+ 
+@@ -1242,7 +1242,7 @@ ztest_dmu_objset_create_destroy(ztest_ar
+ 	if (error) {
+ 		if (error == ENOSPC) {
+ 			ztest_record_enospc("dmu_objset_create");
+-			(void) rw_unlock(&ztest_shared->zs_name_lock);
+			(void) pthread_rwlock_unlock(&ztest_shared->zs_name_lock);
+ 			return;
+ 		}
+ 		fatal(0, "dmu_objset_create(%s) = %d", name, error);
+@@ -1321,7 +1321,7 @@ ztest_dmu_objset_create_destroy(ztest_ar
+ 	if (error)
+ 		fatal(0, "dmu_objset_destroy(%s) = %d", name, error);
+ 
+-	(void) rw_unlock(&ztest_shared->zs_name_lock);
+	(void) pthread_rwlock_unlock(&ztest_shared->zs_name_lock);
+ }
+ 
+ /*
+@@ -1335,7 +1335,7 @@ ztest_dmu_snapshot_create_destroy(ztest_
+ 	char snapname[100];
+ 	char osname[MAXNAMELEN];
+ 
+-	(void) rw_rdlock(&ztest_shared->zs_name_lock);
+	(void) pthread_rwlock_rdlock(&ztest_shared->zs_name_lock);
+ 	dmu_objset_name(os, osname);
+ 	(void) snprintf(snapname, 100, "%s@%llu", osname,
+ 	    (u_longlong_t)za->za_instance);
+@@ -1348,7 +1348,7 @@ ztest_dmu_snapshot_create_destroy(ztest_
+ 		ztest_record_enospc("dmu_take_snapshot");
+ 	else if (error != 0 && error != EEXIST)
+ 		fatal(0, "dmu_take_snapshot() = %d", error);
+-	(void) rw_unlock(&ztest_shared->zs_name_lock);
+	(void) pthread_rwlock_unlock(&ztest_shared->zs_name_lock);
+ }
+ 
+ #define	ZTEST_TRAVERSE_BLOCKS	1000
+@@ -1992,7 +1992,7 @@ ztest_dmu_write_parallel(ztest_args_t *z
+ 	int bs = ZTEST_DIROBJ_BLOCKSIZE;
+ 	int do_free = 0;
+ 	uint64_t off, txg_how;
+-	mutex_t *lp;
+	pthread_mutex_t *lp;
+ 	char osname[MAXNAMELEN];
+ 	char iobuf[SPA_MAXBLOCKSIZE];
+ 	blkptr_t blk = { 0 };
+@@ -2041,7 +2041,7 @@ ztest_dmu_write_parallel(ztest_args_t *z
+ 	}
+ 
+ 	lp = &ztest_shared->zs_sync_lock[b];
+-	(void) mutex_lock(lp);
+	(void) pthread_mutex_lock(lp);
+ 
+ 	wbt->bt_objset = dmu_objset_id(os);
+ 	wbt->bt_object = ZTEST_DIROBJ;
+@@ -2087,7 +2087,7 @@ ztest_dmu_write_parallel(ztest_args_t *z
+ 		dmu_write(os, ZTEST_DIROBJ, off, btsize, wbt, tx);
+ 	}
+ 
+-	(void) mutex_unlock(lp);
+	(void) pthread_mutex_unlock(lp);
+ 
+ 	if (ztest_random(1000) == 0)
+ 		(void) poll(NULL, 0, 1); /* open dn_notxholds window */
+@@ -2106,7 +2106,7 @@ ztest_dmu_write_parallel(ztest_args_t *z
+ 	/*
+ 	 * dmu_sync() the block we just wrote.
+ 	 */
+-	(void) mutex_lock(lp);
+	(void) pthread_mutex_lock(lp);
+ 
+ 	blkoff = P2ALIGN_TYPED(off, bs, uint64_t);
+ 	error = dmu_buf_hold(os, ZTEST_DIROBJ, blkoff, FTAG, &db);
+@@ -2114,7 +2114,7 @@ ztest_dmu_write_parallel(ztest_args_t *z
+ 	if (error) {
+ 		dprintf("dmu_buf_hold(%s, %d, %llx) = %d\n",
+ 		    osname, ZTEST_DIROBJ, blkoff, error);
+-		(void) mutex_unlock(lp);
+		(void) pthread_mutex_unlock(lp);
+ 		return;
+ 	}
+ 	blkoff = off - blkoff;
+@@ -2122,7 +2122,7 @@ ztest_dmu_write_parallel(ztest_args_t *z
+ 	dmu_buf_rele(db, FTAG);
+ 	za->za_dbuf = NULL;
+ 
+-	(void) mutex_unlock(lp);
+	(void) pthread_mutex_unlock(lp);
+ 
+ 	if (error) {
+ 		dprintf("dmu_sync(%s, %d, %llx) = %d\n",
+@@ -2502,7 +2502,7 @@ ztest_dsl_prop_get_set(ztest_args_t *za)
+ 	char osname[MAXNAMELEN];
+ 	int error;
+ 
+-	(void) rw_rdlock(&ztest_shared->zs_name_lock);
+	(void) pthread_rwlock_rdlock(&ztest_shared->zs_name_lock);
+ 
+ 	dmu_objset_name(os, osname);
+ 
+@@ -2541,7 +2541,7 @@ ztest_dsl_prop_get_set(ztest_args_t *za)
+ 		}
+ 	}
+ 
+-	(void) rw_unlock(&ztest_shared->zs_name_lock);
+	(void) pthread_rwlock_unlock(&ztest_shared->zs_name_lock);
+ }
+ 
+ static void
+@@ -2693,7 +2693,7 @@ ztest_spa_rename(ztest_args_t *za)
+ 	int error;
+ 	spa_t *spa;
+ 
+-	(void) rw_wrlock(&ztest_shared->zs_name_lock);
+	(void) pthread_rwlock_wrlock(&ztest_shared->zs_name_lock);
+ 
+ 	oldname = za->za_pool;
+ 	newname = umem_alloc(strlen(oldname) + 5, UMEM_NOFAIL);
+@@ -2745,7 +2745,7 @@ ztest_spa_rename(ztest_args_t *za)
+ 
+ 	umem_free(newname, strlen(newname) + 1);
+ 
+-	(void) rw_unlock(&ztest_shared->zs_name_lock);
+	(void) pthread_rwlock_unlock(&ztest_shared->zs_name_lock);
+ }
+ 
+ 
+@@ -3090,13 +3090,13 @@ ztest_run(char *pool)
+ 	ztest_args_t *za;
+ 	spa_t *spa;
+ 	char name[100];
+-	thread_t tid;
+	pthread_t tid;
+ 
+-	(void) _mutex_init(&zs->zs_vdev_lock, USYNC_THREAD, NULL);
+-	(void) rwlock_init(&zs->zs_name_lock, USYNC_THREAD, NULL);
+	(void) pthread_mutex_init(&zs->zs_vdev_lock, NULL);
+	(void) pthread_rwlock_init(&zs->zs_name_lock, NULL);
+ 
+ 	for (t = 0; t < ZTEST_SYNC_LOCKS; t++)
+-		(void) _mutex_init(&zs->zs_sync_lock[t], USYNC_THREAD, NULL);
+		(void) pthread_mutex_init(&zs->zs_sync_lock[t], NULL);
+ 
+ 	/*
+ 	 * Destroy one disk before we even start.
+@@ -3153,7 +3153,7 @@ ztest_run(char *pool)
+ 	 * start the thread before setting the zio_io_fail_shift, which
+ 	 * will indicate our failure rate.
+ 	 */
+-	error = thr_create(0, 0, ztest_suspend_monitor, NULL, THR_BOUND, &tid);
+	error = pthread_create(&tid, NULL, ztest_suspend_monitor, NULL);
+ 	if (error) {
+ 		fatal(0, "can't create suspend monitor thread: error %d",
+ 		    t, error);
+@@ -3217,7 +3217,7 @@ ztest_run(char *pool)
+ 		if (t < zopt_datasets) {
+ 			ztest_replay_t zr;
+ 			int test_future = FALSE;
+-			(void) rw_rdlock(&ztest_shared->zs_name_lock);
+			(void) pthread_rwlock_rdlock(&ztest_shared->zs_name_lock);
+ 			(void) snprintf(name, 100, "%s/%s_%d", pool, pool, d);
+ 			error = dmu_objset_create(name, DMU_OST_OTHER, NULL, 0,
+ 			    ztest_create_cb, NULL);
+@@ -3225,7 +3225,7 @@ ztest_run(char *pool)
+ 				test_future = TRUE;
+ 			} else if (error == ENOSPC) {
+ 				zs->zs_enospc_count++;
+-				(void) rw_unlock(&ztest_shared->zs_name_lock);
+				(void) pthread_rwlock_unlock(&ztest_shared->zs_name_lock);
+ 				break;
+ 			} else if (error != 0) {
+ 				fatal(0, "dmu_objset_create(%s) = %d",
+@@ -3236,7 +3236,7 @@ ztest_run(char *pool)
+ 			if (error)
+ 				fatal(0, "dmu_objset_open('%s') = %d",
+ 				    name, error);
+-			(void) rw_unlock(&ztest_shared->zs_name_lock);
+			(void) pthread_rwlock_unlock(&ztest_shared->zs_name_lock);
+ 			if (test_future)
+ 				ztest_dmu_check_future_leak(&za[t]);
+ 			zr.zr_os = za[d].za_os;
+@@ -3245,15 +3245,15 @@ ztest_run(char *pool)
+ 			za[d].za_zilog = zil_open(za[d].za_os, NULL);
+ 		}
+ 
+-		error = thr_create(0, 0, ztest_thread, &za[t], THR_BOUND,
+-		    &za[t].za_thread);
+		error = pthread_create(&za[t].za_thread, NULL, ztest_thread,
+		    &za[t]);
+ 		if (error)
+ 			fatal(0, "can't create thread %d: error %d",
+ 			    t, error);
+ 	}
+ 
+ 	while (--t >= 0) {
+-		error = thr_join(za[t].za_thread, NULL, NULL);
+		error = pthread_join(za[t].za_thread, NULL);
+ 		if (error)
+ 			fatal(0, "thr_join(%d) = %d", t, error);
+ 		if (za[t].za_th)
+@@ -3276,14 +3276,14 @@ ztest_run(char *pool)
+ 	 * If we had out-of-space errors, destroy a random objset.
+ 	 */
+ 	if (zs->zs_enospc_count != 0) {
+-		(void) rw_rdlock(&ztest_shared->zs_name_lock);
+		(void) pthread_rwlock_rdlock(&ztest_shared->zs_name_lock);
+ 		d = (int)ztest_random(zopt_datasets);
+ 		(void) snprintf(name, 100, "%s/%s_%d", pool, pool, d);
+ 		if (zopt_verbose >= 3)
+ 			(void) printf("Destroying %s to free up space\n", name);
+ 		(void) dmu_objset_find(name, ztest_destroy_cb, &za[d],
+ 		    DS_FIND_SNAPSHOTS | DS_FIND_CHILDREN);
+-		(void) rw_unlock(&ztest_shared->zs_name_lock);
+		(void) pthread_rwlock_unlock(&ztest_shared->zs_name_lock);
+ 	}
+ 
+ 	txg_wait_synced(spa_get_dsl(spa), 0);
+@@ -3301,7 +3301,7 @@ ztest_run(char *pool)
+ 	mutex_enter(&spa->spa_zio_lock);
+ 	cv_broadcast(&spa->spa_zio_cv);
+ 	mutex_exit(&spa->spa_zio_lock);
+-	error = thr_join(tid, NULL, NULL);
+	error = pthread_join(tid, NULL);
+ 	if (error)
+ 		fatal(0, "thr_join(%d) = %d", tid, error);
+ 
+Index: zfs+chaos4/lib/libuutil/uu_misc.c
+===================================================================
+--- zfs+chaos4.orig/lib/libuutil/uu_misc.c
+++ zfs+chaos4/lib/libuutil/uu_misc.c
+@@ -37,7 +37,6 @@
+ #include <stdlib.h>
+ #include <string.h>
+ #include <sys/debug.h>
+-#include <thread.h>
+ #include <unistd.h>
+ 
+ #if !defined(TEXT_DOMAIN)
+@@ -70,11 +69,12 @@ static va_list		uu_panic_args;
+ static pthread_t	uu_panic_thread;
+ 
+ static uint32_t		_uu_main_error;
+static __thread int	_uu_main_thread = 0;
+ 
+ void
+ uu_set_error(uint_t code)
+ {
+-	if (thr_main() != 0) {
+	if (_uu_main_thread) {
+ 		_uu_main_error = code;
+ 		return;
+ 	}
+@@ -103,7 +103,7 @@ uu_set_error(uint_t code)
+ uint32_t
+ uu_error(void)
+ {
+-	if (thr_main() != 0)
+	if (_uu_main_thread)
+ 		return (_uu_main_error);
+ 
+ 	if (uu_error_key_setup < 0)	/* can't happen? */
+@@ -255,5 +255,6 @@ uu_release_child(void)
+ static void
+ uu_init(void)
+ {
+	_uu_main_thread = 1;
+ 	(void) pthread_atfork(uu_lockup, uu_release, uu_release_child);
+ }
+Index: zfs+chaos4/lib/libzfscommon/include/sys/zfs_context_user.h
+===================================================================
+--- zfs+chaos4.orig/lib/libzfscommon/include/sys/zfs_context_user.h
+++ zfs+chaos4/lib/libzfscommon/include/sys/zfs_context_user.h
+@@ -52,8 +52,7 @@ extern "C" {
+ #include <errno.h>
+ #include <string.h>
+ #include <strings.h>
+-#include <synch.h>
+-#include <thread.h>
+#include <pthread.h>
+ #include <assert.h>
+ #include <alloca.h>
+ #include <umem.h>
+@@ -191,13 +190,15 @@ _NOTE(CONSTCOND) } while (0)
+ /*
+  * Threads
+  */
+-#define	curthread	((void *)(uintptr_t)thr_self())
+
+/* XXX: not portable */
+#define	curthread	((void *)(uintptr_t)pthread_self())
+ 
+ typedef struct kthread kthread_t;
+ 
+ #define	thread_create(stk, stksize, func, arg, len, pp, state, pri)	\
+ 	zk_thread_create(func, arg)
+-#define	thread_exit() thr_exit(NULL)
+#define	thread_exit() pthread_exit(NULL)
+ 
+ extern kthread_t *zk_thread_create(void (*func)(), void *arg);
+ 
+@@ -207,28 +208,18 @@ extern kthread_t *zk_thread_create(void 
+ /*
+  * Mutexes
+  */
+#define MTX_MAGIC 0x9522f51362a6e326ull
+ typedef struct kmutex {
+ 	void		*m_owner;
+-	boolean_t	initialized;
+-	mutex_t		m_lock;
+	uint64_t	m_magic;
+	pthread_mutex_t	m_lock;
+ } kmutex_t;
+ 
+-#define	MUTEX_DEFAULT	USYNC_THREAD
+-#undef MUTEX_HELD
+-#define	MUTEX_HELD(m) _mutex_held(&(m)->m_lock)
+-
+-/*
+- * Argh -- we have to get cheesy here because the kernel and userland
+- * have different signatures for the same routine.
+- */
+-extern int _mutex_init(mutex_t *mp, int type, void *arg);
+-extern int _mutex_destroy(mutex_t *mp);
+-
+-#define	mutex_init(mp, b, c, d)		zmutex_init((kmutex_t *)(mp))
+-#define	mutex_destroy(mp)		zmutex_destroy((kmutex_t *)(mp))
+#define	MUTEX_DEFAULT	0
+#define	MUTEX_HELD(m)	((m)->m_owner == curthread)
+ 
+-extern void zmutex_init(kmutex_t *mp);
+-extern void zmutex_destroy(kmutex_t *mp);
+extern void mutex_init(kmutex_t *mp, char *name, int type, void *cookie);
+extern void mutex_destroy(kmutex_t *mp);
+ extern void mutex_enter(kmutex_t *mp);
+ extern void mutex_exit(kmutex_t *mp);
+ extern int mutex_tryenter(kmutex_t *mp);
+@@ -237,23 +228,24 @@ extern void *mutex_owner(kmutex_t *mp);
+ /*
+  * RW locks
+  */
+#define RW_MAGIC 0x4d31fb123648e78aull
+ typedef struct krwlock {
+-	void		*rw_owner;
+-	boolean_t	initialized;
+-	rwlock_t	rw_lock;
+	void			*rw_owner;
+	void			*rw_wr_owner;
+	uint64_t		rw_magic;
+	pthread_rwlock_t	rw_lock;
+	uint_t			rw_readers;
+ } krwlock_t;
+ 
+ typedef int krw_t;
+ 
+ #define	RW_READER	0
+ #define	RW_WRITER	1
+-#define	RW_DEFAULT	USYNC_THREAD
+-
+-#undef RW_READ_HELD
+-#define	RW_READ_HELD(x)		_rw_read_held(&(x)->rw_lock)
+#define	RW_DEFAULT	0
+ 
+-#undef RW_WRITE_HELD
+-#define	RW_WRITE_HELD(x)	_rw_write_held(&(x)->rw_lock)
+#define	RW_READ_HELD(x)		((x)->rw_readers > 0)
+#define	RW_WRITE_HELD(x)	((x)->rw_wr_owner == curthread)
+#define	RW_LOCK_HELD(x)		(RW_READ_HELD(x) || RW_WRITE_HELD(x))
+ 
+ extern void rw_init(krwlock_t *rwlp, char *name, int type, void *arg);
+ extern void rw_destroy(krwlock_t *rwlp);
+@@ -271,9 +263,13 @@ extern gid_t *crgetgroups(cred_t *cr);
+ /*
+  * Condition variables
+  */
+-typedef cond_t kcondvar_t;
+#define CV_MAGIC 0xd31ea9a83b1b30c4ull
+typedef struct kcondvar {
+	uint64_t cv_magic;
+	pthread_cond_t cv;
+} kcondvar_t;
+ 
+-#define	CV_DEFAULT	USYNC_THREAD
+#define	CV_DEFAULT	0
+ 
+ extern void cv_init(kcondvar_t *cv, char *name, int type, void *arg);
+ extern void cv_destroy(kcondvar_t *cv);
+@@ -444,7 +440,8 @@ extern void delay(clock_t ticks);
+ #define	minclsyspri	60
+ #define	maxclsyspri	99
+ 
+-#define	CPU_SEQID	(thr_self() & (max_ncpus - 1))
+/* XXX: not portable */
+#define	CPU_SEQID	(pthread_self() & (max_ncpus - 1))
+ 
+ #define	kcred		NULL
+ #define	CRED()		NULL
+Index: zfs+chaos4/lib/libzpool/kernel.c
+===================================================================
+--- zfs+chaos4.orig/lib/libzpool/kernel.c
+++ zfs+chaos4/lib/libzpool/kernel.c
+@@ -38,6 +38,7 @@
+ #include <sys/zfs_context.h>
+ #include <sys/zmod.h>
+ #include <sys/utsname.h>
+#include <sys/time.h>
+ 
+ /*
+  * Emulation of kernel services in userland.
+@@ -60,11 +61,15 @@ struct utsname utsname = {
+ kthread_t *
+ zk_thread_create(void (*func)(), void *arg)
+ {
+-	thread_t tid;
+	pthread_t tid;
+ 
+-	VERIFY(thr_create(0, 0, (void *(*)(void *))func, arg, THR_DETACHED,
+-	    &tid) == 0);
+	pthread_attr_t attr;
+	VERIFY(pthread_attr_init(&attr) == 0);
+	VERIFY(pthread_attr_setdetachstate(&attr, PTHREAD_CREATE_DETACHED) == 0);
+ 
+	VERIFY(pthread_create(&tid, &attr, (void *(*)(void *))func, arg) == 0);
+
+	/* XXX: not portable */
+ 	return ((void *)(uintptr_t)tid);
+ }
+ 
+@@ -97,30 +102,37 @@ kstat_delete(kstat_t *ksp)
+  * =========================================================================
+  */
+ void
+-zmutex_init(kmutex_t *mp)
+mutex_init(kmutex_t *mp, char *name, int type, void *cookie)
+ {
+	ASSERT(type == MUTEX_DEFAULT);
+	ASSERT(cookie == NULL);
+
+#ifdef IM_FEELING_LUCKY
+	ASSERT(mp->m_magic != MTX_MAGIC);
+#endif
+
+ 	mp->m_owner = NULL;
+-	mp->initialized = B_TRUE;
+-	(void) _mutex_init(&mp->m_lock, USYNC_THREAD, NULL);
+	mp->m_magic = MTX_MAGIC;
+	VERIFY3S(pthread_mutex_init(&mp->m_lock, NULL), ==, 0);
+ }
+ 
+ void
+-zmutex_destroy(kmutex_t *mp)
+mutex_destroy(kmutex_t *mp)
+ {
+-	ASSERT(mp->initialized == B_TRUE);
+	ASSERT(mp->m_magic == MTX_MAGIC);
+ 	ASSERT(mp->m_owner == NULL);
+-	(void) _mutex_destroy(&(mp)->m_lock);
+	VERIFY3S(pthread_mutex_destroy(&(mp)->m_lock), ==, 0);
+ 	mp->m_owner = (void *)-1UL;
+-	mp->initialized = B_FALSE;
+	mp->m_magic = 0;
+ }
+ 
+ void
+ mutex_enter(kmutex_t *mp)
+ {
+-	ASSERT(mp->initialized == B_TRUE);
+	ASSERT(mp->m_magic == MTX_MAGIC);
+ 	ASSERT(mp->m_owner != (void *)-1UL);
+ 	ASSERT(mp->m_owner != curthread);
+-	VERIFY(mutex_lock(&mp->m_lock) == 0);
+	VERIFY3S(pthread_mutex_lock(&mp->m_lock), ==, 0);
+ 	ASSERT(mp->m_owner == NULL);
+ 	mp->m_owner = curthread;
+ }
+@@ -128,9 +140,9 @@ mutex_enter(kmutex_t *mp)
+ int
+ mutex_tryenter(kmutex_t *mp)
+ {
+-	ASSERT(mp->initialized == B_TRUE);
+	ASSERT(mp->m_magic == MTX_MAGIC);
+ 	ASSERT(mp->m_owner != (void *)-1UL);
+-	if (0 == mutex_trylock(&mp->m_lock)) {
+	if (0 == pthread_mutex_trylock(&mp->m_lock)) {
+ 		ASSERT(mp->m_owner == NULL);
+ 		mp->m_owner = curthread;
+ 		return (1);
+@@ -142,16 +154,16 @@ mutex_tryenter(kmutex_t *mp)
+ void
+ mutex_exit(kmutex_t *mp)
+ {
+-	ASSERT(mp->initialized == B_TRUE);
+	ASSERT(mp->m_magic == MTX_MAGIC);
+ 	ASSERT(mutex_owner(mp) == curthread);
+ 	mp->m_owner = NULL;
+-	VERIFY(mutex_unlock(&mp->m_lock) == 0);
+	VERIFY3S(pthread_mutex_unlock(&mp->m_lock), ==, 0);
+ }
+ 
+ void *
+ mutex_owner(kmutex_t *mp)
+ {
+-	ASSERT(mp->initialized == B_TRUE);
+	ASSERT(mp->m_magic == MTX_MAGIC);
+ 	return (mp->m_owner);
+ }
+ 
+@@ -164,31 +176,48 @@ mutex_owner(kmutex_t *mp)
+ void
+ rw_init(krwlock_t *rwlp, char *name, int type, void *arg)
+ {
+-	rwlock_init(&rwlp->rw_lock, USYNC_THREAD, NULL);
+	ASSERT(type == RW_DEFAULT);
+	ASSERT(arg == NULL);
+
+#ifdef IM_FEELING_LUCKY
+	ASSERT(rwlp->rw_magic != RW_MAGIC);
+#endif
+
+	VERIFY3S(pthread_rwlock_init(&rwlp->rw_lock, NULL), ==, 0);
+ 	rwlp->rw_owner = NULL;
+-	rwlp->initialized = B_TRUE;
+	rwlp->rw_wr_owner = NULL;
+	rwlp->rw_readers = 0;
+	rwlp->rw_magic = RW_MAGIC;
+ }
+ 
+ void
+ rw_destroy(krwlock_t *rwlp)
+ {
+-	rwlock_destroy(&rwlp->rw_lock);
+-	rwlp->rw_owner = (void *)-1UL;
+-	rwlp->initialized = B_FALSE;
+	ASSERT(rwlp->rw_magic == RW_MAGIC);
+
+	VERIFY3S(pthread_rwlock_destroy(&rwlp->rw_lock), ==, 0);
+	rwlp->rw_magic = 0;
+ }
+ 
+ void
+ rw_enter(krwlock_t *rwlp, krw_t rw)
+ {
+-	ASSERT(!RW_LOCK_HELD(rwlp));
+-	ASSERT(rwlp->initialized == B_TRUE);
+-	ASSERT(rwlp->rw_owner != (void *)-1UL);
+	ASSERT(rwlp->rw_magic == RW_MAGIC);
+ 	ASSERT(rwlp->rw_owner != curthread);
+	ASSERT(rwlp->rw_wr_owner != curthread);
+ 
+-	if (rw == RW_READER)
+-		(void) rw_rdlock(&rwlp->rw_lock);
+-	else
+-		(void) rw_wrlock(&rwlp->rw_lock);
+	if (rw == RW_READER) {
+		VERIFY3S(pthread_rwlock_rdlock(&rwlp->rw_lock), ==, 0);
+		ASSERT(rwlp->rw_wr_owner == NULL);
+
+		atomic_inc_uint(&rwlp->rw_readers);
+	} else {
+		VERIFY3S(pthread_rwlock_wrlock(&rwlp->rw_lock), ==, 0);
+		ASSERT(rwlp->rw_wr_owner == NULL);
+		ASSERT3U(rwlp->rw_readers, ==, 0);
+
+		rwlp->rw_wr_owner = curthread;
+	}
+ 
+ 	rwlp->rw_owner = curthread;
+ }
+@@ -196,11 +225,16 @@ rw_enter(krwlock_t *rwlp, krw_t rw)
+ void
+ rw_exit(krwlock_t *rwlp)
+ {
+-	ASSERT(rwlp->initialized == B_TRUE);
+-	ASSERT(rwlp->rw_owner != (void *)-1UL);
+	ASSERT(rwlp->rw_magic == RW_MAGIC);
+	ASSERT(RW_LOCK_HELD(rwlp));
+
+	if (RW_READ_HELD(rwlp))
+		atomic_dec_uint(&rwlp->rw_readers);
+	else
+		rwlp->rw_wr_owner = NULL;
+ 
+ 	rwlp->rw_owner = NULL;
+-	(void) rw_unlock(&rwlp->rw_lock);
+	VERIFY3S(pthread_rwlock_unlock(&rwlp->rw_lock), ==, 0);
+ }
+ 
+ int
+@@ -208,19 +242,29 @@ rw_tryenter(krwlock_t *rwlp, krw_t rw)
+ {
+ 	int rv;
+ 
+-	ASSERT(rwlp->initialized == B_TRUE);
+-	ASSERT(rwlp->rw_owner != (void *)-1UL);
+	ASSERT(rwlp->rw_magic == RW_MAGIC);
+ 
+ 	if (rw == RW_READER)
+-		rv = rw_tryrdlock(&rwlp->rw_lock);
+		rv = pthread_rwlock_tryrdlock(&rwlp->rw_lock);
+ 	else
+-		rv = rw_trywrlock(&rwlp->rw_lock);
+		rv = pthread_rwlock_trywrlock(&rwlp->rw_lock);
+ 
+ 	if (rv == 0) {
+		ASSERT(rwlp->rw_wr_owner == NULL);
+
+		if (rw == RW_READER)
+			atomic_inc_uint(&rwlp->rw_readers);
+		else {
+			ASSERT3U(rwlp->rw_readers, ==, 0);
+			rwlp->rw_wr_owner = curthread;
+		}
+
+ 		rwlp->rw_owner = curthread;
+ 		return (1);
+ 	}
+ 
+	VERIFY3S(rv, ==, EBUSY);
+
+ 	return (0);
+ }
+ 
+@@ -228,8 +272,7 @@ rw_tryenter(krwlock_t *rwlp, krw_t rw)
+ int
+ rw_tryupgrade(krwlock_t *rwlp)
+ {
+-	ASSERT(rwlp->initialized == B_TRUE);
+-	ASSERT(rwlp->rw_owner != (void *)-1UL);
+	ASSERT(rwlp->rw_magic == RW_MAGIC);
+ 
+ 	return (0);
+ }
+@@ -243,22 +286,34 @@ rw_tryupgrade(krwlock_t *rwlp)
+ void
+ cv_init(kcondvar_t *cv, char *name, int type, void *arg)
+ {
+-	VERIFY(cond_init(cv, type, NULL) == 0);
+	ASSERT(type == CV_DEFAULT);
+
+#ifdef IM_FEELING_LUCKY
+	ASSERT(cv->cv_magic != CV_MAGIC);
+#endif
+
+	cv->cv_magic = CV_MAGIC;
+
+	VERIFY3S(pthread_cond_init(&cv->cv, NULL), ==, 0);
+ }
+ 
+ void
+ cv_destroy(kcondvar_t *cv)
+ {
+-	VERIFY(cond_destroy(cv) == 0);
+	ASSERT(cv->cv_magic == CV_MAGIC);
+	VERIFY3S(pthread_cond_destroy(&cv->cv), ==, 0);
+	cv->cv_magic = 0;
+ }
+ 
+ void
+ cv_wait(kcondvar_t *cv, kmutex_t *mp)
+ {
+	ASSERT(cv->cv_magic == CV_MAGIC);
+ 	ASSERT(mutex_owner(mp) == curthread);
+ 	mp->m_owner = NULL;
+-	int ret = cond_wait(cv, &mp->m_lock);
+-	VERIFY(ret == 0 || ret == EINTR);
+	int ret = pthread_cond_wait(&cv->cv, &mp->m_lock);
+	if (ret != 0)
+		VERIFY3S(ret, ==, EINTR);
+ 	mp->m_owner = curthread;
+ }
+ 
+@@ -266,29 +321,38 @@ clock_t
+ cv_timedwait(kcondvar_t *cv, kmutex_t *mp, clock_t abstime)
+ {
+ 	int error;
+	struct timeval tv;
+ 	timestruc_t ts;
+ 	clock_t delta;
+ 
+	ASSERT(cv->cv_magic == CV_MAGIC);
+
+ top:
+ 	delta = abstime - lbolt;
+ 	if (delta <= 0)
+ 		return (-1);
+ 
+-	ts.tv_sec = delta / hz;
+-	ts.tv_nsec = (delta % hz) * (NANOSEC / hz);
+	VERIFY(gettimeofday(&tv, NULL) == 0);
+
+	ts.tv_sec = tv.tv_sec + delta / hz;
+	ts.tv_nsec = tv.tv_usec * 1000 + (delta % hz) * (NANOSEC / hz);
+	if (ts.tv_nsec >= NANOSEC) {
+		ts.tv_sec++;
+		ts.tv_nsec -= NANOSEC;
+	}
+ 
+ 	ASSERT(mutex_owner(mp) == curthread);
+ 	mp->m_owner = NULL;
+-	error = cond_reltimedwait(cv, &mp->m_lock, &ts);
+	error = pthread_cond_timedwait(&cv->cv, &mp->m_lock, &ts);
+ 	mp->m_owner = curthread;
+ 
+-	if (error == ETIME)
+	if (error == ETIMEDOUT)
+ 		return (-1);
+ 
+ 	if (error == EINTR)
+ 		goto top;
+ 
+-	ASSERT(error == 0);
+	VERIFY3S(error, ==, 0);
+ 
+ 	return (1);
+ }
+@@ -296,13 +360,15 @@ top:
+ void
+ cv_signal(kcondvar_t *cv)
+ {
+-	VERIFY(cond_signal(cv) == 0);
+	ASSERT(cv->cv_magic == CV_MAGIC);
+	VERIFY3S(pthread_cond_signal(&cv->cv), ==, 0);
+ }
+ 
+ void
+ cv_broadcast(kcondvar_t *cv)
+ {
+-	VERIFY(cond_broadcast(cv) == 0);
+	ASSERT(cv->cv_magic == CV_MAGIC);
+	VERIFY3S(pthread_cond_broadcast(&cv->cv), ==, 0);
+ }
+ 
+ /*
+@@ -549,11 +615,11 @@ __dprintf(const char *file, const char *
+ 	    dprintf_find_string(func)) {
+ 		/* Print out just the function name if requested */
+ 		flockfile(stdout);
+-		/* XXX: the following printf may not be portable */
+		/* XXX: the following 2 printfs may not be portable */
+ 		if (dprintf_find_string("pid"))
+ 			(void) printf("%llu ", (u_longlong_t) getpid());
+ 		if (dprintf_find_string("tid"))
+-			(void) printf("%u ", (uint_t) thr_self());
+			(void) printf("%u ", (uint_t) pthread_self());
+ 		if (dprintf_find_string("cpu"))
+ 			(void) printf("%u ", getcpuid());
+ 		if (dprintf_find_string("time"))
+Index: zfs+chaos4/lib/libzpool/taskq.c
+===================================================================
+--- zfs+chaos4.orig/lib/libzpool/taskq.c
+++ zfs+chaos4/lib/libzpool/taskq.c
+@@ -43,7 +43,7 @@ struct taskq {
+ 	krwlock_t	tq_threadlock;
+ 	kcondvar_t	tq_dispatch_cv;
+ 	kcondvar_t	tq_wait_cv;
+-	thread_t	*tq_threadlist;
+	pthread_t	*tq_threadlist;
+ 	int		tq_flags;
+ 	int		tq_active;
+ 	int		tq_nthreads;
+@@ -186,7 +186,7 @@ taskq_create(const char *name, int nthre
+ 	tq->tq_maxalloc = maxalloc;
+ 	tq->tq_task.task_next = &tq->tq_task;
+ 	tq->tq_task.task_prev = &tq->tq_task;
+-	tq->tq_threadlist = kmem_alloc(nthreads * sizeof (thread_t), KM_SLEEP);
+	tq->tq_threadlist = kmem_alloc(nthreads * sizeof (pthread_t), KM_SLEEP);
+ 
+ 	if (flags & TASKQ_PREPOPULATE) {
+ 		mutex_enter(&tq->tq_lock);
+@@ -196,8 +196,8 @@ taskq_create(const char *name, int nthre
+ 	}
+ 
+ 	for (t = 0; t < nthreads; t++)
+-		VERIFY(thr_create(0, 0, taskq_thread,
+-		    tq, THR_BOUND, &tq->tq_threadlist[t]) == 0);
+		VERIFY(pthread_create(&tq->tq_threadlist[t],
+		    NULL, taskq_thread, tq) == 0);
+ 
+ 	return (tq);
+ }
+@@ -227,9 +227,9 @@ taskq_destroy(taskq_t *tq)
+ 	mutex_exit(&tq->tq_lock);
+ 
+ 	for (t = 0; t < nthreads; t++)
+-		VERIFY(thr_join(tq->tq_threadlist[t], NULL, NULL) == 0);
+		VERIFY(pthread_join(tq->tq_threadlist[t], NULL) == 0);
+ 
+-	kmem_free(tq->tq_threadlist, nthreads * sizeof (thread_t));
+	kmem_free(tq->tq_threadlist, nthreads * sizeof (pthread_t));
+ 
+ 	rw_destroy(&tq->tq_threadlock);
+ 	mutex_destroy(&tq->tq_lock);
+@@ -248,7 +248,7 @@ taskq_member(taskq_t *tq, void *t)
+ 		return (1);
+ 
+ 	for (i = 0; i < tq->tq_nthreads; i++)
+-		if (tq->tq_threadlist[i] == (thread_t)(uintptr_t)t)
+		if (tq->tq_threadlist[i] == (pthread_t)(uintptr_t)t)
+ 			return (1);
+ 
+ 	return (0);
--- a/patches/zap-cursor-move-to-key.patch
+++ b/patches/zap-cursor-move-to-key.patch
@ -0,0 +1,115 @@
+Add a ZAP API to move a ZAP cursor to a given key.
+
+Index: zfs+chaos4/lib/libzfscommon/include/sys/zap.h
+===================================================================
+--- zfs+chaos4.orig/lib/libzfscommon/include/sys/zap.h
+++ zfs+chaos4/lib/libzfscommon/include/sys/zap.h
+@@ -302,6 +302,11 @@ void zap_cursor_advance(zap_cursor_t *zc
+ uint64_t zap_cursor_serialize(zap_cursor_t *zc);
+ 
+ /*
+ * Advance the cursor to the attribute having the key.
+ */
+int zap_cursor_move_to_key(zap_cursor_t *zc, const char *name, matchtype_t mt);
+
+/*
+  * Initialize a zap cursor pointing to the position recorded by
+  * zap_cursor_serialize (in the "serialized" argument).  You can also
+  * use a "serialized" argument of 0 to start at the beginning of the
+Index: zfs+chaos4/lib/libzfscommon/include/sys/zap_impl.h
+===================================================================
+--- zfs+chaos4.orig/lib/libzfscommon/include/sys/zap_impl.h
+++ zfs+chaos4/lib/libzfscommon/include/sys/zap_impl.h
+@@ -210,6 +210,7 @@ int fzap_add_cd(zap_name_t *zn,
+     uint64_t integer_size, uint64_t num_integers,
+     const void *val, uint32_t cd, dmu_tx_t *tx);
+ void fzap_upgrade(zap_t *zap, dmu_tx_t *tx);
+int fzap_cursor_move_to_key(zap_cursor_t *zc, zap_name_t *zn);
+ 
+ #ifdef	__cplusplus
+ }
+Index: zfs+chaos4/lib/libzpool/zap.c
+===================================================================
+--- zfs+chaos4.orig/lib/libzpool/zap.c
+++ zfs+chaos4/lib/libzpool/zap.c
+@@ -1029,6 +1029,30 @@ zap_stats_ptrtbl(zap_t *zap, uint64_t *t
+ 	}
+ }
+ 
+int fzap_cursor_move_to_key(zap_cursor_t *zc, zap_name_t *zn)
+{
+	int err;
+	zap_leaf_t *l;
+	zap_entry_handle_t zeh;
+	uint64_t hash;
+
+	if (zn->zn_name_orij && strlen(zn->zn_name_orij) > ZAP_MAXNAMELEN)
+		return (E2BIG);
+
+	err = zap_deref_leaf(zc->zc_zap, zn->zn_hash, NULL, RW_READER, &l);
+	if (err != 0)
+		return (err);
+
+	err = zap_leaf_lookup(l, zn, &zeh);
+	if (err != 0)
+		return (err);
+
+	zc->zc_leaf = l;
+	zc->zc_hash = zeh.zeh_hash;
+	zc->zc_cd = zeh.zeh_cd;
+	return 0;
+}
+
+ void
+ fzap_get_stats(zap_t *zap, zap_stats_t *zs)
+ {
+Index: zfs+chaos4/lib/libzpool/zap_micro.c
+===================================================================
+--- zfs+chaos4.orig/lib/libzpool/zap_micro.c
+++ zfs+chaos4/lib/libzpool/zap_micro.c
+@@ -1045,6 +1045,45 @@ zap_cursor_advance(zap_cursor_t *zc)
+ 	}
+ }
+ 
+int zap_cursor_move_to_key(zap_cursor_t *zc, const char *name, matchtype_t mt)
+{
+	int err = 0;
+	mzap_ent_t *mze;
+	zap_name_t *zn;
+
+	if (zc->zc_zap == NULL) {
+		err = zap_lockdir(zc->zc_objset, zc->zc_zapobj, NULL,
+				RW_READER, TRUE, FALSE, &zc->zc_zap);
+		if (err)
+			return (err);
+	} else {
+		rw_enter(&zc->zc_zap->zap_rwlock, RW_READER);
+	}
+
+	zn = zap_name_alloc(zc->zc_zap, name, mt);
+	if (zn == NULL) {
+		rw_exit(&zc->zc_zap->zap_rwlock);
+		return (ENOTSUP);
+	}
+
+	if (!zc->zc_zap->zap_ismicro) {
+		err = fzap_cursor_move_to_key(zc, zn);
+	} else {
+		mze = mze_find(zn);
+		if (mze == NULL) {
+			err = (ENOENT);
+			goto out;
+		}
+		zc->zc_hash = mze->mze_hash;
+		zc->zc_cd = mze->mze_phys.mze_cd;
+	}
+
+out:
+	zap_name_free(zn);
+	rw_exit(&zc->zc_zap->zap_rwlock);
+	return (err);
+}
+
+ int
+ zap_get_stats(objset_t *os, uint64_t zapobj, zap_stats_t *zs)
+ {
--- a/scripts/Makefile.am
+++ b/scripts/Makefile.am
@ -0,0 +1,8 @@
+EXTRA_DIST  = check.sh create-zpool.sh load-zfs.sh unload-zfs.sh
+EXTRA_DIST += profile-kpios-disk.sh profile-kpios-pids.sh
+EXTRA_DIST += profile-kpios-post.sh profile-kpios-pre.sh profile-kpios.sh
+EXTRA_DIST += survey.sh update-zfs.sh zpios-jbod.sh zpios.sh
+
+check:
+	./check.sh
+
--- a/scripts/check.sh
+++ b/scripts/check.sh
@ -0,0 +1,17 @@
+#!/bin/bash
+
+prog=check.sh
+
+die() {
+	echo "${prog}: $1" >&2
+	exit 1
+}
+
+if [ $(id -u) != 0 ]; then
+	die "Must run as root"
+fi
+
+./load-zfs.sh || die ""
+./unload-zfs.sh || die ""
+
+exit 0
--- a/scripts/create-zpool.sh
+++ b/scripts/create-zpool.sh
@ -0,0 +1,42 @@
+#!/bin/bash
+
+prog=create-zpool.sh
+. ../.script-config
+
+# Single disk ilc dev nodes
+DEVICES="/dev/sda"
+
+# All disks in a Thumper config
+#DEVICES="/dev/sda  /dev/sdb  /dev/sdc  /dev/sdd  /dev/sde  /dev/sdf  \
+#         /dev/sdg  /dev/sdh  /dev/sdi  /dev/sdj  /dev/sdk  /dev/sdl  \
+#         /dev/sdm  /dev/sdn  /dev/sdo  /dev/sdp  /dev/sdq  /dev/sdr  \
+#         /dev/sds  /dev/sdt  /dev/sdu  /dev/sdv  /dev/sdw  /dev/sdx  \
+#         /dev/sdy  /dev/sdz  /dev/sdaa /dev/sdab /dev/sdac /dev/sdad \
+#         /dev/sdae /dev/sdaf /dev/sdag /dev/sdah /dev/sdai /dev/sdaj \
+#         /dev/sdak /dev/sdal /dev/sdam /dev/sdan /dev/sdao /dev/sdap \
+#         /dev/sdaq /dev/sdar /dev/sdas /dev/sdat /dev/sdau /dev/sdav"
+
+# Sun style disk in Thumper config
+#DEVICES="/dev/sda  /dev/sdb  /dev/sdc \
+#         /dev/sdi  /dev/sdj  /dev/sdk \
+#         /dev/sdr  /dev/sds  /dev/sdt \
+#         /dev/sdz  /dev/sdaa /dev/sdab"
+
+# Promise JBOD config (ilc23)
+#DEVICES="/dev/sdb  /dev/sdc  /dev/sdd \
+#         /dev/sde  /dev/sdf  /dev/sdg \
+#         /dev/sdh  /dev/sdi  /dev/sdj \
+#         /dev/sdk  /dev/sdl  /dev/sdm"
+
+echo
+echo "zpool create lustre <devices>"
+${CMDDIR}/zpool/zpool create -F lustre ${DEVICES}
+
+echo
+echo "zpool list"
+${CMDDIR}/zpool/zpool list
+
+echo
+echo "zpool status lustre"
+${CMDDIR}/zpool/zpool status lustre
+
--- a/scripts/load-zfs.sh
+++ b/scripts/load-zfs.sh
@ -0,0 +1,58 @@
+#!/bin/bash
+
+prog=load-zfs.sh
+. ../.script-config
+
+spl_options=$1
+zpool_options=$2
+
+spl_module=${SPLBUILD}/modules/spl/spl.ko
+zlib_module=/lib/modules/${KERNELSRCVER}/kernel/lib/zlib_deflate/zlib_deflate.ko
+zavl_module=${ZFSBUILD}/lib/libavl/zavl.ko
+znvpair_module=${ZFSBUILD}/lib/libnvpair/znvpair.ko
+zport_module=${ZFSBUILD}/lib/libport/zport.ko
+zcommon_module=${ZFSBUILD}/lib/libzcommon/zcommon.ko
+zpool_module=${ZFSBUILD}/lib/libzpool/zpool.ko
+zctl_module=${ZFSBUILD}/lib/libdmu-ctl/zctl.ko
+zpios_module=${ZFSBUILD}/lib/libzpios/zpios.ko
+
+die() {
+	echo "${prog}: $1" >&2
+	exit 1
+}
+
+load_module() {
+	echo "Loading $1"
+	/sbin/insmod $* || die "Failed to load $1"
+}
+
+if [ $(id -u) != 0 ]; then
+	die "Must run as root"
+fi
+
+if /sbin/lsmod | egrep -q "^spl|^zavl|^znvpair|^zport|^zcommon|^zlib_deflate|^zpool"; then
+	die "Must start with modules unloaded"
+fi
+
+if [ ! -f ${zavl_module} ] ||
+   [ ! -f ${znvpair_module} ] ||
+   [ ! -f ${zport_module} ] ||
+   [ ! -f ${zcommon_module} ] ||
+   [ ! -f ${zpool_module} ]; then
+	die "Source tree must be built, run 'make'"
+fi
+
+load_module ${spl_module} ${spl_options}
+load_module ${zlib_module}
+load_module ${zavl_module}
+load_module ${znvpair_module}
+load_module ${zport_module}
+load_module ${zcommon_module}
+load_module ${zpool_module} ${zpool_options}
+load_module ${zctl_module}
+load_module ${zpios_module}
+
+sleep 1
+echo "Successfully loaded ZFS module stack"
+
+exit 0
--- a/scripts/profile-kpios-disk.sh
+++ b/scripts/profile-kpios-disk.sh
@ -0,0 +1,128 @@
+#!/bin/bash
+# profile-kpios-disk.sh
+#
+# /proc/diskinfo <after skipping major/minor>
+# Field 1 -- device name
+# Field 2 -- # of reads issued
+# Field 3 -- # of reads merged
+# Field 4 -- # of sectors read
+# Field 5 -- # of milliseconds spent reading
+# Field 6 -- # of writes completed
+# Field 7 -- # of writes merged
+# Field 8 -- # of sectors written
+# Field 9 -- # of milliseconds spent writing
+# Field 10 -- # of I/Os currently in progress
+# Field 11 -- # of milliseconds spent doing I/Os
+# Field 12 -- weighted # of milliseconds spent doing I/Os
+
+RUN_PIDS=${0}
+RUN_LOG_DIR=${1}
+RUN_ID=${2}
+
+create_table() {
+	local FIELD=$1
+	local ROW_M=()
+	local ROW_N=()
+	local HEADER=1
+	local STEP=1
+
+	for DISK_FILE in `ls -r --sort=time --time=ctime ${RUN_LOG_DIR}/${RUN_ID}/disk-[0-9]*`; do
+		ROW_M=( ${ROW_N[@]} )
+		ROW_N=( `cat ${DISK_FILE} | grep sd | cut -c11- | cut -f${FIELD} -d' ' | tr "\n" "\t"` )
+
+		if [ $HEADER -eq 1 ]; then
+			echo -n "step, "
+			cat ${DISK_FILE} | grep sd | cut -c11- | cut -f1 -d' ' | tr "\n" ", "
+	                echo "total"
+			HEADER=0
+		fi
+
+		if [ ${#ROW_M[@]} -eq 0 ]; then
+			continue
+		fi
+
+		if [ ${#ROW_M[@]} -ne ${#ROW_N[@]} ]; then
+			echo "Badly formatted profile data in ${DISK_FILE}"
+			break
+		fi
+
+		TOTAL=0
+		echo -n "${STEP}, "
+		for (( i=0; i<${#ROW_N[@]}; i++ )); do
+			DELTA=`echo "${ROW_N[${i}]}-${ROW_M[${i}]}" | bc`
+			let TOTAL=${TOTAL}+${DELTA}
+			echo -n "${DELTA}, "
+		done
+		echo "${TOTAL}, "
+
+		let STEP=${STEP}+1
+	done
+}
+
+create_table_mbs() {
+	local FIELD=$1
+	local TIME=$2
+	local ROW_M=()
+	local ROW_N=()
+	local HEADER=1
+	local STEP=1
+
+	for DISK_FILE in `ls -r --sort=time --time=ctime ${RUN_LOG_DIR}/${RUN_ID}/disk-[0-9]*`; do
+		ROW_M=( ${ROW_N[@]} )
+		ROW_N=( `cat ${DISK_FILE} | grep sd | cut -c11- | cut -f${FIELD} -d' ' | tr "\n" "\t"` )
+
+		if [ $HEADER -eq 1 ]; then
+			echo -n "step, "
+			cat ${DISK_FILE} | grep sd | cut -c11- | cut -f1 -d' ' | tr "\n" ", "
+	                echo "total"
+			HEADER=0
+		fi
+
+		if [ ${#ROW_M[@]} -eq 0 ]; then
+			continue
+		fi
+
+		if [ ${#ROW_M[@]} -ne ${#ROW_N[@]} ]; then
+			echo "Badly formatted profile data in ${DISK_FILE}"
+			break
+		fi
+
+		TOTAL=0
+		echo -n "${STEP}, "
+		for (( i=0; i<${#ROW_N[@]}; i++ )); do
+			DELTA=`echo "${ROW_N[${i}]}-${ROW_M[${i}]}" | bc`
+			MBS=`echo "scale=2; ((${DELTA}*512)/${TIME})/(1024*1024)" | bc`
+			TOTAL=`echo "scale=2; ${TOTAL}+${MBS}" | bc`
+			echo -n "${MBS}, "
+		done
+		echo "${TOTAL}, "
+
+		let STEP=${STEP}+1
+	done
+}
+
+echo
+echo "Reads issued per device"
+create_table 2
+echo
+echo "Reads merged per device"
+create_table 3
+echo
+echo "Sectors read per device"
+create_table 4
+echo "MB/s per device"
+create_table_mbs 4 3
+
+echo
+echo "Writes issued per device"
+create_table 6
+echo
+echo "Writes merged per device"
+create_table 7
+echo
+echo "Sectors written per device"
+create_table 8
+echo "MB/s per device"
+create_table_mbs 8 3
+
+exit 0
--- a/scripts/profile-kpios-pids.sh
+++ b/scripts/profile-kpios-pids.sh
@ -0,0 +1,130 @@
+#!/bin/bash
+# profile-kpios-pids.sh
+
+RUN_PIDS=${0}
+RUN_LOG_DIR=${1}
+RUN_ID=${2}
+
+ROW_M=()
+ROW_N=()
+ROW_N_SCHED=()
+ROW_N_WAIT=()
+
+HEADER=1
+STEP=1
+
+for PID_FILE in `ls -r --sort=time --time=ctime ${RUN_LOG_DIR}/${RUN_ID}/pids-[0-9]*`; do
+	ROW_M=( ${ROW_N[@]} )
+	ROW_N=( 0 0 0 0  0 0 0 0  0 0 0 0  0 0 0 0  0 0 0 )
+	ROW_N_SCHED=( `cat ${PID_FILE} | cut -f15 -d' ' | tr "\n" "\t"` )
+	ROW_N_WAIT=(  `cat ${PID_FILE} | cut -f17 -d' ' | tr "\n" "\t"` )
+	ROW_N_NAMES=( `cat ${PID_FILE} | cut -f2  -d' ' | cut -f2 -d'(' | 
+                       cut -f1 -d')'   | cut -f1  -d'/' | tr "\n" "\t"` )
+
+	for (( i=0; i<${#ROW_N_SCHED[@]}; i++ )); do
+		SUM=`echo "${ROW_N_WAIT[${i}]}+${ROW_N_SCHED[${i}]}" | bc`
+
+		case ${ROW_N_NAMES[${i}]} in
+			zio_taskq)	IDX=0;;
+			zio_req_nul)	IDX=1;;
+			zio_irq_nul)	IDX=2;;
+			zio_req_rd)	IDX=3;;
+			zio_irq_rd)	IDX=4;;
+			zio_req_wr)	IDX=5;;
+			zio_irq_wr)	IDX=6;;
+			zio_req_fr)	IDX=7;;
+			zio_irq_fr)	IDX=8;;
+			zio_req_cm)	IDX=9;;
+			zio_irq_cm)	IDX=10;;
+			zio_req_ctl)	IDX=11;;
+			zio_irq_ctl)	IDX=12;;
+			txg_quiesce)	IDX=13;;
+			txg_sync)	IDX=14;;
+			txg_timelimit)	IDX=15;;
+			arc_reclaim)	IDX=16;;
+			l2arc_feed)	IDX=17;;
+			kpios_io)	IDX=18;;
+			*)		continue;;
+		esac
+
+		let ROW_N[${IDX}]=${ROW_N[${IDX}]}+${SUM}
+	done
+
+	if [ $HEADER -eq 1 ]; then
+		echo "step, zio_taskq, zio_req_nul, zio_irq_nul, "        \
+                     "zio_req_rd, zio_irq_rd, zio_req_wr, zio_irq_wr, "   \
+                     "zio_req_fr, zio_irq_fr, zio_req_cm, zio_irq_cm, "   \
+                     "zio_req_ctl, zio_irq_ctl, txg_quiesce, txg_sync, "  \
+                     "txg_timelimit, arc_reclaim, l2arc_feed, kpios_io, " \
+		     "idle"
+		HEADER=0
+	fi
+
+	if [ ${#ROW_M[@]} -eq 0 ]; then
+		continue
+	fi
+
+	if [ ${#ROW_M[@]} -ne ${#ROW_N[@]} ]; then
+		echo "Badly formatted profile data in ${PID_FILE}"
+		break
+	fi
+
+	# Original values are in jiffies and we expect HZ to be 1000
+	# on most 2.6 systems thus we divide by 10 to get a percentage.
+	IDLE=1000
+        echo -n "${STEP}, "
+	for (( i=0; i<${#ROW_N[@]}; i++ )); do
+		DELTA=`echo "${ROW_N[${i}]}-${ROW_M[${i}]}" | bc`
+		DELTA_PERCENT=`echo "scale=1; ${DELTA}/10" | bc`
+		let IDLE=${IDLE}-${DELTA}
+		echo -n "${DELTA_PERCENT}, "
+	done
+	ILDE_PERCENT=`echo "scale=1; ${IDLE}/10" | bc`
+	echo "${ILDE_PERCENT}"
+
+	let STEP=${STEP}+1
+done
+
+exit
+
+echo
+echo "Percent of total system time per pid"
+for PID_FILE in `ls -r --sort=time --time=ctime ${RUN_LOG_DIR}/${RUN_ID}/pids-[0-9]*`; do
+	ROW_M=( ${ROW_N[@]} )
+	ROW_N_SCHED=( `cat ${PID_FILE} | cut -f15 -d' ' | tr "\n" "\t"` )
+	ROW_N_WAIT=( `cat ${PID_FILE} | cut -f17 -d' ' | tr "\n" "\t"` )
+
+	for (( i=0; i<${#ROW_N_SCHED[@]}; i++ )); do
+		ROW_N[${i}]=`echo "${ROW_N_WAIT[${i}]}+${ROW_N_SCHED[${i}]}" | bc`
+	done
+
+	if [ $HEADER -eq 1 ]; then
+		echo -n "step, "
+		cat ${PID_FILE} | cut -f2 -d' ' | tr "\n" ", "
+		echo
+		HEADER=0
+	fi
+
+	if [ ${#ROW_M[@]} -eq 0 ]; then
+		continue
+	fi
+
+	if [ ${#ROW_M[@]} -ne ${#ROW_N[@]} ]; then
+		echo "Badly formatted profile data in ${PID_FILE}"
+		break
+	fi
+
+	# Original values are in jiffies and we expect HZ to be 1000
+	# on most 2.6 systems thus we divide by 10 to get a percentage.
+        echo -n "${STEP}, "
+	for (( i=0; i<${#ROW_N[@]}; i++ )); do
+		DELTA=`echo "scale=1; (${ROW_N[${i}]}-${ROW_M[${i}]})/10" | bc`
+		echo -n "${DELTA}, "
+	done
+
+	echo
+	let STEP=${STEP}+1
+done
+
+
+exit 0
--- a/scripts/profile-kpios-post.sh
+++ b/scripts/profile-kpios-post.sh
@ -0,0 +1,67 @@
+#!/bin/bash
+
+prog=profile-kpios-post.sh
+. ../.script-config
+
+RUN_POST=${0}
+RUN_PHASE=${1}
+RUN_LOG_DIR=${2}
+RUN_ID=${3}
+RUN_POOL=${4}
+RUN_CHUNK_SIZE=${5}
+RUN_REGION_SIZE=${6}
+RUN_THREAD_COUNT=${7}
+RUN_REGION_COUNT=${8}
+RUN_OFFSET=${9}
+RUN_REGION_NOISE=${10}
+RUN_CHUNK_NOISE=${11}
+RUN_THREAD_DELAY=${12}
+RUN_FLAGS=${13}
+RUN_RESULT=${14}
+
+PROFILE_KPIOS_PIDS_BIN=/home/behlendo/src/zfs/scripts/profile-kpios-pids.sh
+PROFILE_KPIOS_PIDS_LOG=${RUN_LOG_DIR}/${RUN_ID}/pids-summary.csv
+
+PROFILE_KPIOS_DISK_BIN=/home/behlendo/src/zfs/scripts/profile-kpios-disk.sh
+PROFILE_KPIOS_DISK_LOG=${RUN_LOG_DIR}/${RUN_ID}/disk-summary.csv
+
+PROFILE_KPIOS_ARC_LOG=${RUN_LOG_DIR}/${RUN_ID}/arcstats
+PROFILE_KPIOS_VDEV_LOG=${RUN_LOG_DIR}/${RUN_ID}/vdev_cache_stats
+
+KERNEL_BIN="/lib/modules/`uname -r`/kernel/"
+SPL_BIN="${SPLBUILD}/modules/spl/"
+ZFS_BIN="${ZFSBUILD}/lib/"
+
+OPROFILE_SHORT_ARGS="-a -g -l -p ${KERNEL_BIN},${SPL_BIN},${ZFS_BIN}"
+OPROFILE_LONG_ARGS="-d -a -g -l -p ${KERNEL_BIN},${SPL_BIN},${ZFS_BIN}"
+
+OPROFILE_LOG=${RUN_LOG_DIR}/${RUN_ID}/oprofile.txt
+OPROFILE_SHORT_LOG=${RUN_LOG_DIR}/${RUN_ID}/oprofile-short.txt
+OPROFILE_LONG_LOG=${RUN_LOG_DIR}/${RUN_ID}/oprofile-long.txt
+PROFILE_PID=${RUN_LOG_DIR}/${RUN_ID}/pid
+
+if [ "${RUN_PHASE}" != "post" ]; then
+        exit 1
+fi
+
+# opcontrol --stop >>${OPROFILE_LOG} 2>&1
+# opcontrol --dump >>${OPROFILE_LOG} 2>&1
+
+kill -s SIGHUP `cat ${PROFILE_PID}`
+rm -f ${PROFILE_PID}
+
+# opreport ${OPROFILE_SHORT_ARGS} >${OPROFILE_SHORT_LOG} 2>&1
+# opreport ${OPROFILE_LONG_ARGS} >${OPROFILE_LONG_LOG} 2>&1
+
+# opcontrol --deinit >>${OPROFILE_LOG} 2>&1
+
+cat /proc/spl/kstat/zfs/arcstats >${PROFILE_KPIOS_ARC_LOG}
+cat /proc/spl/kstat/zfs/vdev_cache_stats >${PROFILE_KPIOS_VDEV_LOG}
+
+# Summarize system time per pid
+${PROFILE_KPIOS_PIDS_BIN} ${RUN_LOG_DIR} ${RUN_ID} >${PROFILE_KPIOS_PIDS_LOG}
+
+# Summarize per device performance
+${PROFILE_KPIOS_DISK_BIN} ${RUN_LOG_DIR} ${RUN_ID} >${PROFILE_KPIOS_DISK_LOG}
+
+exit 0
--- a/scripts/profile-kpios-pre.sh
+++ b/scripts/profile-kpios-pre.sh
@ -0,0 +1,69 @@
+#!/bin/bash
+# profile-kpios-pre.sh
+
+trap "PROFILE_KPIOS_READY=1" SIGHUP
+
+RUN_PRE=${0}
+RUN_PHASE=${1}
+RUN_LOG_DIR=${2}
+RUN_ID=${3}
+RUN_POOL=${4}
+RUN_CHUNK_SIZE=${5}
+RUN_REGION_SIZE=${6}
+RUN_THREAD_COUNT=${7}
+RUN_REGION_COUNT=${8}
+RUN_OFFSET=${9}
+RUN_REGION_NOISE=${10}
+RUN_CHUNK_NOISE=${11}
+RUN_THREAD_DELAY=${12}
+RUN_FLAGS=${13}
+RUN_RESULT=${14}
+
+PROFILE_KPIOS_BIN=/home/behlendo/src/zfs/scripts/profile-kpios.sh
+PROFILE_KPIOS_READY=0
+
+OPROFILE_LOG=${RUN_LOG_DIR}/${RUN_ID}/oprofile.txt
+PROFILE_PID=${RUN_LOG_DIR}/${RUN_ID}/pid
+RUN_ARGS=${RUN_LOG_DIR}/${RUN_ID}/args
+
+if [ "${RUN_PHASE}" != "pre" ]; then
+        exit 1
+fi
+
+rm -Rf ${RUN_LOG_DIR}/${RUN_ID}/
+mkdir -p ${RUN_LOG_DIR}/${RUN_ID}/
+
+echo "PHASE=${RUN_PHASE}" >>${RUN_ARGS}
+echo "LOG_DIR=${RUN_LOG_DIR}" >>${RUN_ARGS}
+echo "ID=${RUN_ID}" >>${RUN_ARGS}
+echo "POOL=${RUN_POOL}" >>${RUN_ARGS}
+echo "CHUNK_SIZE=${RUN_CHUNK_SIZE}" >>${RUN_ARGS}
+echo "REGION_SIZE=${RUN_REGION_SIZE}" >>${RUN_ARGS}
+echo "THREAD_COUNT=${RUN_THREAD_COUNT}" >>${RUN_ARGS}
+echo "REGION_COUNT=${RUN_REGION_COUNT}" >>${RUN_ARGS}
+echo "OFFSET=${RUN_OFFSET}" >>${RUN_ARGS}
+echo "REGION_NOISE=${RUN_REGION_NOISE}" >>${RUN_ARGS}
+echo "CHUNK_NOISE=${RUN_CHUNK_NOISE}" >>${RUN_ARGS}
+echo "THREAD_DELAY=${RUN_THREAD_DELAY}" >>${RUN_ARGS}
+echo "FLAGS=${RUN_FLAGS}" >>${RUN_ARGS}
+echo "RESULT=${RUN_RESULT}" >>${RUN_ARGS}
+
+# XXX: Oprofile support seems to be broken when I try and start
+# it via a user mode helper script, I suspect the setup is failing.
+# opcontrol --init >>${OPROFILE_LOG} 2>&1
+# opcontrol --setup --vmlinux=/boot/vmlinux >>${OPROFILE_LOG} 2>&1
+
+# Start the profile script
+${PROFILE_KPIOS_BIN} ${RUN_PHASE} ${RUN_LOG_DIR} ${RUN_ID} &
+echo "$!" >${PROFILE_PID}
+
+# Sleep waiting for profile script to be ready, it will
+# signal us via SIGHUP when it is ready to start profiling.
+while [ ${PROFILE_KPIOS_READY} -eq 0 ]; do
+	sleep 0.1
+done
+
+# opcontrol --start-daemon >>${OPROFILE_LOG} 2>&1
+# opcontrol --start >>${OPROFILE_LOG} 2>&1
+
+exit 0
--- a/scripts/profile-kpios.sh
+++ b/scripts/profile-kpios.sh
@ -0,0 +1,222 @@
+#!/bin/bash
+# profile-kpios.sh
+
+trap "RUN_DONE=1" SIGHUP
+
+RUN_PHASE=${1}
+RUN_LOG_DIR=${2}
+RUN_ID=${3}
+RUN_DONE=0
+
+POLL_INTERVAL=2.99
+
+# Log these pids, the exact pid numbers will vary from system to system
+# so I harvest pid for all the following type of processes from /proc/<pid>/
+#
+# zio_taskq/#
+# spa_zio_issue/#
+# spa_zio_intr/#
+# txg_quiesce_thr
+# txg_sync_thread
+# txg_timelimit_t
+# arc_reclaim_thr
+# l2arc_feed_thre
+# kpios_io/#
+
+ZIO_TASKQ_PIDS=()
+ZIO_REQ_NUL_PIDS=() 
+ZIO_IRQ_NUL_PIDS=() 
+ZIO_REQ_RD_PIDS=() 
+ZIO_IRQ_RD_PIDS=() 
+ZIO_REQ_WR_PIDS=() 
+ZIO_IRQ_WR_PIDS=()
+ZIO_REQ_FR_PIDS=() 
+ZIO_IRQ_FR_PIDS=()
+ZIO_REQ_CM_PIDS=() 
+ZIO_IRQ_CM_PIDS=()
+ZIO_REQ_CTL_PIDS=() 
+ZIO_IRQ_CTL_PIDS=()
+
+TXG_QUIESCE_PIDS=()
+TXG_SYNC_PIDS=() 
+TXG_TIMELIMIT_PIDS=()
+
+ARC_RECLAIM_PIDS=()
+L2ARC_FEED_PIDS=()
+
+KPIOS_IO_PIDS=()
+
+show_pids() {
+	echo "* zio_taskq:     { ${ZIO_TASKQ_PIDS[@]} } = ${#ZIO_TASKQ_PIDS[@]}"
+	echo "* zio_req_nul:   { ${ZIO_REQ_NUL_PIDS[@]} } = ${#ZIO_REQ_NUL_PIDS[@]}"
+	echo "* zio_irq_nul:   { ${ZIO_IRQ_NUL_PIDS[@]} } = ${#ZIO_IRQ_NUL_PIDS[@]}"
+	echo "* zio_req_rd:    { ${ZIO_REQ_RD_PIDS[@]} } = ${#ZIO_REQ_RD_PIDS[@]}"
+	echo "* zio_irq_rd:    { ${ZIO_IRQ_RD_PIDS[@]} } = ${#ZIO_IRQ_RD_PIDS[@]}"
+	echo "* zio_req_wr:    { ${ZIO_REQ_WR_PIDS[@]} } = ${#ZIO_REQ_WR_PIDS[@]}"
+	echo "* zio_irq_wr:    { ${ZIO_IRQ_WR_PIDS[@]} } = ${#ZIO_IRQ_WR_PIDS[@]}"
+	echo "* zio_req_fr:    { ${ZIO_REQ_FR_PIDS[@]} } = ${#ZIO_REQ_FR_PIDS[@]}"
+	echo "* zio_irq_fr:    { ${ZIO_IRQ_FR_PIDS[@]} } = ${#ZIO_IRQ_FR_PIDS[@]}"
+	echo "* zio_req_cm:    { ${ZIO_REQ_CM_PIDS[@]} } = ${#ZIO_REQ_CM_PIDS[@]}"
+	echo "* zio_irq_cm:    { ${ZIO_IRQ_CM_PIDS[@]} } = ${#ZIO_IRQ_CM_PIDS[@]}"
+	echo "* zio_req_ctl:   { ${ZIO_REQ_CTL_PIDS[@]} } = ${#ZIO_REQ_CTL_PIDS[@]}"
+	echo "* zio_irq_ctl:   { ${ZIO_IRQ_CTL_PIDS[@]} } = ${#ZIO_IRQ_CTL_PIDS[@]}"
+	echo "* txg_quiesce:   { ${TXG_QUIESCE_PIDS[@]} } = ${#TXG_QUIESCE_PIDS[@]}"
+	echo "* txg_sync:      { ${TXG_SYNC_PIDS[@]} } = ${#TXG_SYNC_PIDS[@]}"
+	echo "* txg_timelimit: { ${TXG_TIMELIMIT_PIDS[@]} } = ${#TXG_TIMELIMIT_PIDS[@]}"
+	echo "* arc_reclaim:   { ${ARC_RECLAIM_PIDS[@]} } = ${#ARC_RECLAIM_PIDS[@]}"
+	echo "* l2arc_feed:    { ${L2ARC_FEED_PIDS[@]} } = ${#L2ARC_FEED_PIDS[@]}"
+	echo "* kpios_io:      { ${KPIOS_IO_PIDS[@]} } = ${#KPIOS_IO_PIDS[@]}"
+}
+
+check_pid() {
+	local PID=$1
+	local NAME=$2
+	local TYPE=$3
+	local PIDS=( "$4" )
+        local NAME_STRING=`echo ${NAME} | cut -f1 -d'/'`
+        local NAME_NUMBER=`echo ${NAME} | cut -f2 -d'/'`
+
+	if [ "${NAME_STRING}" == "${TYPE}" ]; then
+		if [ -n "${NAME_NUMBER}" ]; then
+			PIDS[${NAME_NUMBER}]=${PID}
+		else
+			PIDS[${#PIDS[@]}]=${PID}
+
+		fi
+	fi
+
+	echo "${PIDS[@]}"
+}
+
+# NOTE: This whole process is crazy slow but it will do for now
+aquire_pids() {
+	echo "--- Aquiring ZFS pids ---"
+
+	for PID in `ls /proc/ | grep [0-9] | sort -n -u`; do
+		if [ ! -e /proc/${PID}/status ]; then
+			continue
+		fi
+
+		NAME=`cat /proc/${PID}/status  | head -n1 | cut -f2`
+
+		ZIO_TASKQ_PIDS=( `check_pid ${PID} ${NAME} "zio_taskq" \
+		                 "$(echo "${ZIO_TASKQ_PIDS[@]}")"` )
+
+		ZIO_REQ_NUL_PIDS=( `check_pid ${PID} ${NAME} "zio_req_nul" \
+		                   "$(echo "${ZIO_REQ_NUL_PIDS[@]}")"` )
+
+		ZIO_IRQ_NUL_PIDS=( `check_pid ${PID} ${NAME} "zio_irq_nul" \
+		                   "$(echo "${ZIO_IRQ_NUL_PIDS[@]}")"` )
+
+		ZIO_REQ_RD_PIDS=( `check_pid ${PID} ${NAME} "zio_req_rd" \
+		                   "$(echo "${ZIO_REQ_RD_PIDS[@]}")"` )
+
+		ZIO_IRQ_RD_PIDS=( `check_pid ${PID} ${NAME} "zio_irq_rd" \
+		                   "$(echo "${ZIO_IRQ_RD_PIDS[@]}")"` )
+
+		ZIO_REQ_WR_PIDS=( `check_pid ${PID} ${NAME} "zio_req_wr" \
+		                   "$(echo "${ZIO_REQ_WR_PIDS[@]}")"` )
+
+		ZIO_IRQ_WR_PIDS=( `check_pid ${PID} ${NAME} "zio_irq_wr" \
+		                   "$(echo "${ZIO_IRQ_WR_PIDS[@]}")"` )
+
+		ZIO_REQ_FR_PIDS=( `check_pid ${PID} ${NAME} "zio_req_fr" \
+		                   "$(echo "${ZIO_REQ_FR_PIDS[@]}")"` )
+
+		ZIO_IRQ_FR_PIDS=( `check_pid ${PID} ${NAME} "zio_irq_fr" \
+		                   "$(echo "${ZIO_IRQ_FR_PIDS[@]}")"` )
+
+		ZIO_REQ_CM_PIDS=( `check_pid ${PID} ${NAME} "zio_req_cm" \
+		                   "$(echo "${ZIO_REQ_CM_PIDS[@]}")"` )
+
+		ZIO_IRQ_CM_PIDS=( `check_pid ${PID} ${NAME} "zio_irq_cm" \
+		                   "$(echo "${ZIO_IRQ_CM_PIDS[@]}")"` )
+
+		ZIO_REQ_CTL_PIDS=( `check_pid ${PID} ${NAME} "zio_req_ctl" \
+		                   "$(echo "${ZIO_REQ_CTL_PIDS[@]}")"` )
+
+		ZIO_IRQ_CTL_PIDS=( `check_pid ${PID} ${NAME} "zio_irq_ctl" \
+		                   "$(echo "${ZIO_IRQ_CTL_PIDS[@]}")"` )
+
+		TXG_QUIESCE_PIDS=( `check_pid ${PID} ${NAME} "txg_quiesce" \
+		                   "$(echo "${TXG_QUIESCE_PIDS[@]}")"` )
+
+		TXG_SYNC_PIDS=( `check_pid ${PID} ${NAME} "txg_sync" \
+		                "$(echo "${TXG_SYNC_PIDS[@]}")"` )
+
+		TXG_TIMELIMIT_PIDS=( `check_pid ${PID} ${NAME} "txg_timelimit" \
+		                     "$(echo "${TXG_TIMELIMIT_PIDS[@]}")"` )
+
+		ARC_RECLAIM_PIDS=( `check_pid ${PID} ${NAME} "arc_reclaim" \
+                                     "$(echo "${ARC_RECLAIM_PIDS[@]}")"` )
+
+		L2ARC_FEED_PIDS=( `check_pid ${PID} ${NAME} "l2arc_feed" \
+                                  "$(echo "${L2ARC_FEED_PIDS[@]}")"` )
+	done
+
+	# Wait for kpios_io threads to start
+	kill -s SIGHUP ${PPID}
+	echo "* Waiting for kpios_io threads to start"
+	while [ ${RUN_DONE} -eq 0 ]; do
+		KPIOS_IO_PIDS=( `ps ax | grep kpios_io | grep -v grep | \
+                                 sed 's/^ *//g' | cut -f1 -d' '` )
+		if [ ${#KPIOS_IO_PIDS[@]} -gt 0 ]; then
+			break;
+		fi
+		sleep 0.1
+	done
+
+	echo "`show_pids`" >${RUN_LOG_DIR}/${RUN_ID}/pids.txt
+}
+
+log_pids() {
+	echo "--- Logging ZFS profile to ${RUN_LOG_DIR}/${RUN_ID}/ ---"
+	ALL_PIDS=( ${ZIO_TASKQ_PIDS[@]}     \
+                   ${ZIO_REQ_NUL_PIDS[@]}   \
+                   ${ZIO_IRQ_NUL_PIDS[@]}   \
+                   ${ZIO_REQ_RD_PID[@]}     \
+                   ${ZIO_IRQ_RD_PIDS[@]}    \
+                   ${ZIO_REQ_WR_PIDS[@]}    \
+                   ${ZIO_IRQ_WR_PIDS[@]}    \
+                   ${ZIO_REQ_FR_PIDS[@]}    \ 
+                   ${ZIO_IRQ_FR_PIDS[@]}    \
+                   ${ZIO_REQ_CM_PIDS[@]}    \ 
+                   ${ZIO_IRQ_CM_PIDS[@]}    \
+                   ${ZIO_REQ_CTL_PIDS[@]}   \
+                   ${ZIO_IRQ_CTL_PIDS[@]}   \
+                   ${TXG_QUIESCE_PIDS[@]}   \
+                   ${TXG_SYNC_PIDS[@]}      \
+                   ${TXG_TIMELIMIT_PIDS[@]} \
+                   ${ARC_RECLAIM_PIDS[@]}   \
+                   ${L2ARC_FEED_PIDS[@]}    \
+                   ${KPIOS_IO_PIDS[@]} )
+
+	while [ ${RUN_DONE} -eq 0 ]; do
+		NOW=`date +%s.%N`
+		LOG_PIDS="${RUN_LOG_DIR}/${RUN_ID}/pids-${NOW}"
+		LOG_DISK="${RUN_LOG_DIR}/${RUN_ID}/disk-${NOW}"
+
+		for PID in "${ALL_PIDS[@]}"; do
+			if [ -z ${PID} ]; then
+				continue;
+			fi
+
+	        	if [ -e /proc/${PID}/stat ]; then
+	        		cat /proc/${PID}/stat | head -n1 >>${LOG_PIDS}
+			else
+	                	echo "<${PID} exited>" >>${LOG_PIDS}
+	        	fi
+		done
+
+		cat /proc/diskstats >${LOG_DISK}
+
+		NOW2=`date +%s.%N`
+		DELTA=`echo "${POLL_INTERVAL}-(${NOW2}-${NOW})" | bc`
+		sleep ${DELTA}
+	done
+}
+
+aquire_pids
+log_pids
+
+exit 0
--- a/scripts/survey.sh
+++ b/scripts/survey.sh
@ -0,0 +1,102 @@
+#!/bin/bash
+
+prog=survey.sh
+. ../.script-config
+
+LOG=/home/`whoami`/zpios-logs/`uname -r`/kpios-`date +%Y%m%d`/
+mkdir -p ${LOG}
+
+# Apply all tunings described below to generate some best case
+# numbers for what is acheivable with some more elbow grease.
+NAME="prefetch+zerocopy+checksum+pending1024+kmem"
+echo "----------------------- ${NAME} ------------------------------"
+./zpios.sh                                                           \
+	""                                                           \
+	"zfs_prefetch_disable=1 zfs_vdev_max_pending=1024 zio_bulk_flags=0x100" \
+        "--zerocopy"                                                 \
+	${LOG}/${NAME}/                                              \
+        "${CMDDIR}/zfs/zfs set checksum=off lustre" |       \
+	tee ${LOG}/${NAME}.txt
+
+# Baseline number for an out of the box config with no manual tuning.
+# Ideally, we will want things to be automatically tuned and for this
+# number to approach the tweaked out results above.
+NAME="baseline"
+echo "----------------------- ${NAME} ------------------------------"
+./zpios.sh                                                           \
+	""                                                           \
+	""                                                           \
+        ""                                                           \
+	${LOG}/${NAME}/ |                                            \
+	tee ${LOG}/${NAME}.txt
+
+# Disable ZFS's prefetching.  For some reason still not clear to me
+# current prefetching policy is quite bad for a random workload.
+# Allow the algorithm to detect a random workload and not do anything
+# may be the way to address this issue.
+NAME="prefetch"
+echo "----------------------- ${NAME} ------------------------------"
+./zpios.sh                                                           \
+	""                                                           \
+	"zfs_prefetch_disable=1"                                     \
+        ""                                                           \
+	${LOG}/${NAME}/ |                                            \
+	tee ${LOG}/${NAME}.txt
+
+# As expected, simulating a zerocopy IO path improves performance
+# by freeing up lots of CPU which is wasted move data between buffers.
+NAME="zerocopy"
+echo "----------------------- ${NAME} ------------------------------"
+./zpios.sh                                                           \
+	""                                                           \
+	""                                                           \
+        "--zerocopy"                                                 \
+	${LOG}/${NAME}/ |                                            \
+	tee ${LOG}/${NAME}.txt
+
+# Disabling checksumming should show some (if small) improvement
+# simply due to freeing up a modest amount of CPU.
+NAME="checksum"
+echo "----------------------- ${NAME} ------------------------------"
+./zpios.sh                                                           \
+	""                                                           \
+	""                                                           \
+        ""                                                           \
+	${LOG}/${NAME}/                                              \
+        "${CMDDIR}/zfs/zfs set checksum=off lustre" |       \
+	tee ${LOG}/${NAME}.txt
+
+# Increasing the pending IO depth also seems to improve things likely
+# at the expense of latency.  This should be exported more because I'm
+# seeing a much bigger impact there that I would have expected.  There
+# may be some low hanging fruit to be found here.
+NAME="pending"
+echo "----------------------- ${NAME} ------------------------------"
+./zpios.sh                                                           \
+	""                                                           \
+	"zfs_vdev_max_pending=1024"                                  \
+        ""                                                           \
+	${LOG}/${NAME}/ |                                            \
+	tee ${LOG}/${NAME}.txt
+
+# To avoid memory fragmentation issues our slab implementation can be
+# based on a virtual address space.  Interestingly, we take a pretty
+# substantial performance penalty for this somewhere in the low level
+# IO drivers.  If we back the slab with kmem pages we see far better
+# read performance numbers at the cost of memory fragmention and general
+# system instability due to large allocations.  This may be because of
+# an optimization in the low level drivers due to the contigeous kmem
+# based memory.  This needs to be explained.  The good news here is that
+# with zerocopy interfaces added at the DMU layer we could gaurentee
+# kmem based memory for a pool of pages.
+#
+# 0x100 = KMC_KMEM - Force kmem_* based slab
+# 0x200 = KMC_VMEM - Force vmem_* based slab
+NAME="kmem"
+echo "----------------------- ${NAME} ------------------------------"
+./zpios.sh                                                           \
+	""                                                           \
+	"zio_bulk_flags=0x100"                                       \
+        ""                                                           \
+	${LOG}/${NAME}/ |                                            \
+	tee ${LOG}/${NAME}.txt
--- a/scripts/unload-zfs.sh
+++ b/scripts/unload-zfs.sh
@ -0,0 +1,55 @@
+#!/bin/bash
+
+prog=unload-zfs.sh
+. ../.script-config
+
+spl_module=${SPLBUILD}/modules/spl/spl.ko
+zlib_module=/lib/modules/${KERNELSRCVER}/kernel/lib/zlib_deflate/zlib_deflate.ko
+zavl_module=${ZFSBUILD}/lib/libavl/zavl.ko
+znvpair_module=${ZFSBUILD}/lib/libnvpair/znvpair.ko
+zport_module=${ZFSBUILD}/lib/libport/zport.ko
+zcommon_module=${ZFSBUILD}/lib/libzcommon/zcommon.ko
+zpool_module=${ZFSBUILD}/lib/libzpool/zpool.ko
+zctl_module=${ZFSBUILD}/lib/libdmu-ctl/zctl.ko
+zpios_module=${ZFSBUILD}/lib/libzpios/zpios.ko
+
+die() {
+	echo "${prog}: $1" >&2
+	exit 1
+}
+
+unload_module() {
+	echo "Unloading $1"
+	/sbin/rmmod $1 || die "Failed to unload $1"
+}
+
+if [ $(id -u) != 0 ]; then
+	die "Must run as root"
+fi
+
+unload_module ${zpios_module}
+unload_module ${zctl_module}
+unload_module ${zpool_module}
+unload_module ${zcommon_module}
+unload_module ${zport_module}
+unload_module ${znvpair_module}
+unload_module ${zavl_module}
+unload_module ${zlib_module}
+
+# Set DUMP=1 to generate debug logs on unload
+if [ -n "${DUMP}" ]; then
+	sysctl -w kernel.spl.debug.dump=1
+	# This is racy, I don't like it, but for a helper script it will do.
+	SPL_LOG=`dmesg | tail -n 1 | cut -f5 -d' '`
+	${SPLBUILD}/cmd/spl ${SPL_LOG} >${SPL_LOG}.log
+	echo
+	echo "Dumped debug log: ${SPL_LOG}.log"
+	tail -n1 ${SPL_LOG}.log
+	echo
+fi
+
+unload_module ${spl_module}
+
+echo "Successfully unloaded ZFS module stack"
+
+exit 0
--- a/scripts/update-zfs.sh
+++ b/scripts/update-zfs.sh
@ -0,0 +1,59 @@
+#!/bin/bash
+
+PROG=update-zfs.sh
+ZFS_SRC=http://dlc.sun.com/osol/on/downloads/b89/on-src.tar.bz2
+
+die() {
+	rm -Rf $SRC
+	echo "${PROG}: $1" >&2
+	exit 1
+}
+
+DEST=`pwd`
+if [ `basename $DEST` != "scripts" ]; then
+	die "Must be run from scripts directory"
+fi
+
+SRC=`mktemp -d /tmp/zfs.XXXXXXXXXX`
+DEST=`dirname $DEST`
+DATE=`date +%Y%m%d%H%M%S`
+
+wget $ZFS_SRC
+
+echo "--- Updating ZFS source ---"
+echo
+echo "ZFS_REPO       = $ZFS_REPO"
+echo "ZFS_PATCH_REPO = $ZFS_PATCH_REPO"
+echo "SRC            = $SRC"
+echo "DEST           = $DEST"
+
+echo
+echo "--- Cloning $ZFS_REPO ---"
+cd $SRC || die "Failed to 'cd $SRC'"
+hg clone $ZFS_REPO || die "Failed to clone $ZFS_REPO"
+
+echo
+echo "--- Cloning $ZFS_PATCH_REPO ---"
+hg clone $ZFS_PATCH_REPO patches || die "Failed to clone $ZFS_PATCH_REPO"
+
+echo
+echo "--- Backing up existing files ---"
+echo "$DEST/zfs -> $DEST/zfs.$DATE"
+cp -Rf $DEST/zfs $DEST/zfs.$DATE || die "Failed to backup"
+echo "$DEST/zfs_patches -> $DEST/zfs_patches.$DATE"
+cp -Rf $DEST/zfs_patches $DEST/zfs_patches.$DATE || die "Failed to backup"
+
+echo
+echo "--- Overwriting $DEST/zfs and $DEST/zfs_patches ---"
+find $SRC/trunk/src/ -name SConstruct -type f -print | xargs /bin/rm -f
+find $SRC/trunk/src/ -name SConscript -type f -print | xargs /bin/rm -f
+find $SRC/trunk/src/ -name *.orig -type f -print | xargs /bin/rm -f
+rm -f $SRC/trunk/src/myconfig.py
+cp -Rf $SRC/trunk/src/* $DEST/zfs || die "Failed to overwrite"
+cp -Rf $SRC/patches/*.patch $DEST/zfs_patches/patches/ || die "Failed to overwrite"
+cp -f $SRC/patches/series $DEST/zfs_patches/series/zfs-lustre
+
+echo
+echo "--- Removing $SRC ---"
+rm -Rf $SRC
+
--- a/scripts/zpios-jbod.sh
+++ b/scripts/zpios-jbod.sh
@ -0,0 +1,110 @@
+#!/bin/bash
+
+prog=zpios-jbod.sh
+. ../.script-config
+
+SPL_OPTIONS=$1
+ZPOOL_OPTIONS=$2
+KPIOS_OPTIONS=$3
+PROFILE_KPIOS_LOGS=$4
+KPIOS_PRE=$5
+KPIOS_POST=$6
+
+PROFILE_KPIOS_PRE=/home/behlendo/src/zfs/scripts/profile-kpios-pre.sh
+PROFILE_KPIOS_POST=/home/behlendo/src/zfs/scripts/profile-kpios-post.sh
+
+echo ------------------------- ZFS TEST LOG ---------------------------------
+echo -n "Date = "; date
+echo -n "Kernel = "; uname -r
+echo ------------------------------------------------------------------------
+
+echo
+./load-zfs.sh "${SPL_OPTIONS}" "${ZPOOL_OPTIONS}"
+
+sysctl -w kernel.spl.debug.mask=0
+sysctl -w kernel.spl.debug.subsystem=0
+
+echo ---------------------- SPL Sysctl Tunings ------------------------------
+sysctl -A | grep spl
+echo
+
+echo ------------------- SPL/ZPOOL Module Tunings ---------------------------
+grep [0-9] /sys/module/spl/parameters/*
+grep [0-9] /sys/module/zpool/parameters/*
+echo
+
+DEVICES="/dev/sdn  /dev/sdo  /dev/sdp \
+         /dev/sdq  /dev/sdr  /dev/sds \
+         /dev/sdt  /dev/sdu  /dev/sdv \
+         /dev/sdw  /dev/sdx  /dev/sdy"
+
+${CMDDIR}/zpool/zpool create -F lustre ${DEVICES}
+${CMDDIR}/zpool/zpool status lustre
+
+if [ -n "${KPIOS_PRE}" ]; then
+	${KPIOS_PRE}
+fi 
+
+# Usage: zpios
+#         --chunksize         -c    =values
+#         --chunksize_low     -a    =value
+#         --chunksize_high    -b    =value
+#         --chunksize_incr    -g    =value
+#         --offset            -o    =values
+#         --offset_low        -m    =value
+#         --offset_high       -q    =value
+#         --offset_incr       -r    =value
+#         --regioncount       -n    =values
+#         --regioncount_low   -i    =value
+#         --regioncount_high  -j    =value
+#         --regioncount_incr  -k    =value
+#         --threadcount       -t    =values
+#         --threadcount_low   -l    =value
+#         --threadcount_high  -h    =value
+#         --threadcount_incr  -e    =value
+#         --regionsize        -s    =values
+#         --regionsize_low    -A    =value
+#         --regionsize_high   -B    =value
+#         --regionsize_incr   -C    =value
+#         --cleanup           -x
+#         --verify            -V
+#         --zerocopy          -z
+#         --threaddelay       -T    =jiffies
+#         --regionnoise       -I    =shift
+#         --chunknoise        -N    =bytes
+#         --prerun            -P    =pre-command
+#         --postrun           -R    =post-command
+#         --log               -G    =log directory
+#         --pool | --path     -p    =pool name
+#         --load              -L    =dmuio
+#         --help              -?    =this help
+#         --verbose           -v    =increase verbosity
+#	--threadcount=256,256,256,256,256                        \
+
+CMD="${CMDDIR}/zpios/zpios                                       \
+	--load=dmuio                                             \
+	--path=lustre                                            \
+	--chunksize=1M                                           \
+	--regionsize=4M                                          \
+	--regioncount=16384                                      \
+	--threadcount=256                        \
+	--offset=4M                                              \
+        --cleanup                                                \
+	--verbose                                                \
+	--human-readable                                         \
+	${KPIOS_OPTIONS}                                         \
+        --prerun=${PROFILE_KPIOS_PRE}                            \
+        --postrun=${PROFILE_KPIOS_POST}                          \
+        --log=${PROFILE_KPIOS_LOGS}" 
+echo
+date
+echo ${CMD}
+$CMD
+date
+
+if [ -n "${KPIOS_POST}" ]; then
+	${KPIOS_POST}
+fi 
+
+${CMDDIR}/zpool/zpool destroy lustre
+./unload-zfs.sh
--- a/scripts/zpios.sh
+++ b/scripts/zpios.sh
@ -0,0 +1,139 @@
+#!/bin/bash
+
+prog=zpios.sh
+. ../.script-config
+
+SPL_OPTIONS="spl_debug_mask=0 spl_debug_subsys=0 ${1}"
+ZPOOL_OPTIONS=$2
+KPIOS_OPTIONS=$3
+PROFILE_KPIOS_LOGS=$4
+KPIOS_PRE=$5
+KPIOS_POST=$6
+
+PROFILE_KPIOS_PRE=/home/behlendo/src/zfs/scripts/profile-kpios-pre.sh
+PROFILE_KPIOS_POST=/home/behlendo/src/zfs/scripts/profile-kpios-post.sh
+
+echo ------------------------- ZFS TEST LOG ---------------------------------
+echo -n "Date = "; date
+echo -n "Kernel = "; uname -r
+echo ------------------------------------------------------------------------
+
+echo
+./load-zfs.sh "${SPL_OPTIONS}" "${ZPOOL_OPTIONS}"
+
+echo ---------------------- SPL Sysctl Tunings ------------------------------
+sysctl -A | grep spl
+echo
+
+echo ------------------- SPL/ZPOOL Module Tunings ---------------------------
+if [ -d /sys/module/spl/parameters ]; then
+	grep [0-9] /sys/module/spl/parameters/*
+	grep [0-9] /sys/module/zpool/parameters/*
+else
+	grep [0-9] /sys/module/spl/*
+	grep [0-9] /sys/module/zpool/*
+fi
+echo
+
+# LOCAL HACK
+if [ `hostname` = "ilc23" ]; then
+	DEVICES="/dev/sdy  /dev/sdo  /dev/sdp  /dev/sdq  /dev/sdr  /dev/sds \
+	         /dev/sdt  /dev/sdu  /dev/sdv  /dev/sdw  /dev/sdx"
+else
+	DEVICES="/dev/hda"
+fi
+
+echo "${CMDDIR}/zpool/zpool create -F lustre ${DEVICES}"
+${CMDDIR}/zpool/zpool create -F lustre ${DEVICES}
+
+echo "${CMDDIR}/zpool/zpool status lustre"
+${CMDDIR}/zpool/zpool status lustre
+
+echo "Waiting for /dev/kpios to come up..."
+while [ ! -c /dev/kpios ]; do
+	sleep 1
+done
+
+if [ -n "${KPIOS_PRE}" ]; then
+	${KPIOS_PRE}
+fi 
+
+# Usage: zpios
+#         --chunksize         -c    =values
+#         --chunksize_low     -a    =value
+#         --chunksize_high    -b    =value
+#         --chunksize_incr    -g    =value
+#         --offset            -o    =values
+#         --offset_low        -m    =value
+#         --offset_high       -q    =value
+#         --offset_incr       -r    =value
+#         --regioncount       -n    =values
+#         --regioncount_low   -i    =value
+#         --regioncount_high  -j    =value
+#         --regioncount_incr  -k    =value
+#         --threadcount       -t    =values
+#         --threadcount_low   -l    =value
+#         --threadcount_high  -h    =value
+#         --threadcount_incr  -e    =value
+#         --regionsize        -s    =values
+#         --regionsize_low    -A    =value
+#         --regionsize_high   -B    =value
+#         --regionsize_incr   -C    =value
+#         --cleanup           -x
+#         --verify            -V
+#         --zerocopy          -z
+#         --threaddelay       -T    =jiffies
+#         --regionnoise       -I    =shift
+#         --chunknoise        -N    =bytes
+#         --prerun            -P    =pre-command
+#         --postrun           -R    =post-command
+#         --log               -G    =log directory
+#         --pool | --path     -p    =pool name
+#         --load              -L    =dmuio
+#         --help              -?    =this help
+#         --verbose           -v    =increase verbosity
+
+#        --prerun=${PROFILE_KPIOS_PRE}                            \
+#        --postrun=${PROFILE_KPIOS_POST}                          \
+
+CMD="${CMDDIR}/zpios/zpios                                       \
+	--load=dmuio                                             \
+	--path=lustre                                            \
+	--chunksize=1M                                           \
+	--regionsize=4M                                          \
+	--regioncount=16384                                      \
+	--threadcount=256,256,256,256,256                        \
+	--offset=4M                                              \
+        --cleanup                                                \
+	--verbose                                                \
+	--human-readable                                         \
+	${KPIOS_OPTIONS}                                         \
+        --log=${PROFILE_KPIOS_LOGS}" 
+echo
+date
+echo ${CMD}
+$CMD
+date
+
+if [ -n "${KPIOS_POST}" ]; then
+	${KPIOS_POST}
+fi 
+
+${CMDDIR}/zpool/zpool destroy lustre
+
+echo ---------------------- SPL Sysctl Tunings ------------------------------
+sysctl -A | grep spl
+echo
+
+echo ------------------------ KSTAT Statistics ------------------------------
+echo ARCSTATS
+cat /proc/spl/kstat/zfs/arcstats
+echo
+echo VDEV_CACHE_STATS
+cat /proc/spl/kstat/zfs/vdev_cache_stats
+echo
+echo SLAB
+cat /proc/spl/kmem/slab
+echo
+
+./unload-zfs.sh
--- a/zfs/Makefile.in
+++ b/zfs/Makefile.in
@ -0,0 +1,16 @@
+subdir-m += lib
+subdir-m += zcmd
+
+all:
+# Make the exported SPL symbols available to this module.  There
+# is probably a better way to do this, but this will have to do
+# for now...  an option to modpost perhaps.
+	cp @splsymvers@ .
+
+# Kick off the kernel build system
+	$(MAKE) -C @LINUX@ SUBDIRS=`pwd` @KERNELMAKE_PARAMS@ modules
+
+install uninstall clean distclean maintainer-clean distdir:
+	$(MAKE) -C @LINUX@ SUBDIRS=`pwd` @KERNELMAKE_PARAMS@ $@
+
+check:
--- a/zfs/lib/Makefile.in
+++ b/zfs/lib/Makefile.in
@ -0,0 +1,12 @@
+subdir-m += libuutil		# User space util support
+subdir-m += libumem		# User space memory support
+subdir-m += libzfs		# User space library support
+subdir-m += libsolcompat	# User space compatibility library
+
+subdir-m += libzpool		# Kernel DMU/SPA
+subdir-m += libdmu-ctl		# Kernel control interface
+
+subdir-m += libavl		# Kernel + user space AVL tree support
+subdir-m += libnvpair		# Kernel + user space name/value support
+subdir-m += libzcommon		# Kernel + user space common support
+subdir-m += libport		# Kernel + user space linux support
--- a/zfs/lib/libavl/Makefile.in
+++ b/zfs/lib/libavl/Makefile.in
@ -0,0 +1,31 @@
+subdir-m += include
+DISTFILES = avl.c
+
+MODULE  := zavl
+LIBRARY := libavl
+
+# Compile as kernel module.  Needed symlinks created for all
+# k* objects created by top level configure script.
+
+EXTRA_CFLAGS  = @KERNELCPPFLAGS@
+EXTRA_CFLAGS += -I@LIBDIR@/libavl/include
+
+obj-m := ${MODULE}.o
+
+${MODULE}-objs += kavl.o	# Generic AVL support
+
+# Compile as shared library.  There's an extra useless host program
+# here called 'zu' because it was the easiest way I could convince
+# the kernel build system to construct a user space shared library.
+
+HOSTCFLAGS += @HOSTCFLAGS@
+HOSTCFLAGS += -I@LIBDIR@/libsolcompat/include
+HOSTCFLAGS += -I@LIBDIR@/libport/include
+HOSTCFLAGS += -I@LIBDIR@/libavl/include
+
+hostprogs-y := zu
+always := $(hostprogs-y)
+
+zu-objs := zu.o ${LIBRARY}.so
+
+${LIBRARY}-objs += uavl.o
--- a/zfs/lib/libavl/avl.c
+++ b/zfs/lib/libavl/avl.c
@ -0,0 +1,969 @@
+/*
+ * CDDL HEADER START
+ *
+ * The contents of this file are subject to the terms of the
+ * Common Development and Distribution License (the "License").
+ * You may not use this file except in compliance with the License.
+ *
+ * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
+ * or http://www.opensolaris.org/os/licensing.
+ * See the License for the specific language governing permissions
+ * and limitations under the License.
+ *
+ * When distributing Covered Code, include this CDDL HEADER in each
+ * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
+ * If applicable, add the following below this CDDL HEADER, with the
+ * fields enclosed by brackets "[]" replaced with your own identifying
+ * information: Portions Copyright [yyyy] [name of copyright owner]
+ *
+ * CDDL HEADER END
+ */
+/*
+ * Copyright 2006 Sun Microsystems, Inc.  All rights reserved.
+ * Use is subject to license terms.
+ */
+
+
+
+
+/*
+ * AVL - generic AVL tree implementation for kernel use
+ *
+ * A complete description of AVL trees can be found in many CS textbooks.
+ *
+ * Here is a very brief overview. An AVL tree is a binary search tree that is
+ * almost perfectly balanced. By "almost" perfectly balanced, we mean that at
+ * any given node, the left and right subtrees are allowed to differ in height
+ * by at most 1 level.
+ *
+ * This relaxation from a perfectly balanced binary tree allows doing
+ * insertion and deletion relatively efficiently. Searching the tree is
+ * still a fast operation, roughly O(log(N)).
+ *
+ * The key to insertion and deletion is a set of tree maniuplations called
+ * rotations, which bring unbalanced subtrees back into the semi-balanced state.
+ *
+ * This implementation of AVL trees has the following peculiarities:
+ *
+ *	- The AVL specific data structures are physically embedded as fields
+ *	  in the "using" data structures.  To maintain generality the code
+ *	  must constantly translate between "avl_node_t *" and containing
+ *	  data structure "void *"s by adding/subracting the avl_offset.
+ *
+ *	- Since the AVL data is always embedded in other structures, there is
+ *	  no locking or memory allocation in the AVL routines. This must be
+ *	  provided for by the enclosing data structure's semantics. Typically,
+ *	  avl_insert()/_add()/_remove()/avl_insert_here() require some kind of
+ *	  exclusive write lock. Other operations require a read lock.
+ *
+ *      - The implementation uses iteration instead of explicit recursion,
+ *	  since it is intended to run on limited size kernel stacks. Since
+ *	  there is no recursion stack present to move "up" in the tree,
+ *	  there is an explicit "parent" link in the avl_node_t.
+ *
+ *      - The left/right children pointers of a node are in an array.
+ *	  In the code, variables (instead of constants) are used to represent
+ *	  left and right indices.  The implementation is written as if it only
+ *	  dealt with left handed manipulations.  By changing the value assigned
+ *	  to "left", the code also works for right handed trees.  The
+ *	  following variables/terms are frequently used:
+ *
+ *		int left;	// 0 when dealing with left children,
+ *				// 1 for dealing with right children
+ *
+ *		int left_heavy;	// -1 when left subtree is taller at some node,
+ *				// +1 when right subtree is taller
+ *
+ *		int right;	// will be the opposite of left (0 or 1)
+ *		int right_heavy;// will be the opposite of left_heavy (-1 or 1)
+ *
+ *		int direction;  // 0 for "<" (ie. left child); 1 for ">" (right)
+ *
+ *	  Though it is a little more confusing to read the code, the approach
+ *	  allows using half as much code (and hence cache footprint) for tree
+ *	  manipulations and eliminates many conditional branches.
+ *
+ *	- The avl_index_t is an opaque "cookie" used to find nodes at or
+ *	  adjacent to where a new value would be inserted in the tree. The value
+ *	  is a modified "avl_node_t *".  The bottom bit (normally 0 for a
+ *	  pointer) is set to indicate if that the new node has a value greater
+ *	  than the value of the indicated "avl_node_t *".
+ */
+
+#include <sys/types.h>
+#include <sys/param.h>
+#include <sys/debug.h>
+#include <sys/avl.h>
+#include <sys/cmn_err.h>
+
+/*
+ * Small arrays to translate between balance (or diff) values and child indeces.
+ *
+ * Code that deals with binary tree data structures will randomly use
+ * left and right children when examining a tree.  C "if()" statements
+ * which evaluate randomly suffer from very poor hardware branch prediction.
+ * In this code we avoid some of the branch mispredictions by using the
+ * following translation arrays. They replace random branches with an
+ * additional memory reference. Since the translation arrays are both very
+ * small the data should remain efficiently in cache.
+ */
+static const int  avl_child2balance[2]	= {-1, 1};
+static const int  avl_balance2child[]	= {0, 0, 1};
+
+
+/*
+ * Walk from one node to the previous valued node (ie. an infix walk
+ * towards the left). At any given node we do one of 2 things:
+ *
+ * - If there is a left child, go to it, then to it's rightmost descendant.
+ *
+ * - otherwise we return thru parent nodes until we've come from a right child.
+ *
+ * Return Value:
+ * NULL - if at the end of the nodes
+ * otherwise next node
+ */
+void *
+avl_walk(avl_tree_t *tree, void	*oldnode, int left)
+{
+	size_t off = tree->avl_offset;
+	avl_node_t *node = AVL_DATA2NODE(oldnode, off);
+	int right = 1 - left;
+	int was_child;
+
+
+	/*
+	 * nowhere to walk to if tree is empty
+	 */
+	if (node == NULL)
+		return (NULL);
+
+	/*
+	 * Visit the previous valued node. There are two possibilities:
+	 *
+	 * If this node has a left child, go down one left, then all
+	 * the way right.
+	 */
+	if (node->avl_child[left] != NULL) {
+		for (node = node->avl_child[left];
+		    node->avl_child[right] != NULL;
+		    node = node->avl_child[right])
+			;
+	/*
+	 * Otherwise, return thru left children as far as we can.
+	 */
+	} else {
+		for (;;) {
+			was_child = AVL_XCHILD(node);
+			node = AVL_XPARENT(node);
+			if (node == NULL)
+				return (NULL);
+			if (was_child == right)
+				break;
+		}
+	}
+
+	return (AVL_NODE2DATA(node, off));
+}
+
+/*
+ * Return the lowest valued node in a tree or NULL.
+ * (leftmost child from root of tree)
+ */
+void *
+avl_first(avl_tree_t *tree)
+{
+	avl_node_t *node;
+	avl_node_t *prev = NULL;
+	size_t off = tree->avl_offset;
+
+	for (node = tree->avl_root; node != NULL; node = node->avl_child[0])
+		prev = node;
+
+	if (prev != NULL)
+		return (AVL_NODE2DATA(prev, off));
+	return (NULL);
+}
+
+/*
+ * Return the highest valued node in a tree or NULL.
+ * (rightmost child from root of tree)
+ */
+void *
+avl_last(avl_tree_t *tree)
+{
+	avl_node_t *node;
+	avl_node_t *prev = NULL;
+	size_t off = tree->avl_offset;
+
+	for (node = tree->avl_root; node != NULL; node = node->avl_child[1])
+		prev = node;
+
+	if (prev != NULL)
+		return (AVL_NODE2DATA(prev, off));
+	return (NULL);
+}
+
+/*
+ * Access the node immediately before or after an insertion point.
+ *
+ * "avl_index_t" is a (avl_node_t *) with the bottom bit indicating a child
+ *
+ * Return value:
+ *	NULL: no node in the given direction
+ *	"void *"  of the found tree node
+ */
+void *
+avl_nearest(avl_tree_t *tree, avl_index_t where, int direction)
+{
+	int child = AVL_INDEX2CHILD(where);
+	avl_node_t *node = AVL_INDEX2NODE(where);
+	void *data;
+	size_t off = tree->avl_offset;
+
+	if (node == NULL) {
+		ASSERT(tree->avl_root == NULL);
+		return (NULL);
+	}
+	data = AVL_NODE2DATA(node, off);
+	if (child != direction)
+		return (data);
+
+	return (avl_walk(tree, data, direction));
+}
+
+
+/*
+ * Search for the node which contains "value".  The algorithm is a
+ * simple binary tree search.
+ *
+ * return value:
+ *	NULL: the value is not in the AVL tree
+ *		*where (if not NULL)  is set to indicate the insertion point
+ *	"void *"  of the found tree node
+ */
+void *
+avl_find(avl_tree_t *tree, void *value, avl_index_t *where)
+{
+	avl_node_t *node;
+	avl_node_t *prev = NULL;
+	int child = 0;
+	int diff;
+	size_t off = tree->avl_offset;
+
+	for (node = tree->avl_root; node != NULL;
+	    node = node->avl_child[child]) {
+
+		prev = node;
+
+		diff = tree->avl_compar(value, AVL_NODE2DATA(node, off));
+		ASSERT(-1 <= diff && diff <= 1);
+		if (diff == 0) {
+#ifdef DEBUG
+			if (where != NULL)
+				*where = 0;
+#endif
+			return (AVL_NODE2DATA(node, off));
+		}
+		child = avl_balance2child[1 + diff];
+
+	}
+
+	if (where != NULL)
+		*where = AVL_MKINDEX(prev, child);
+
+	return (NULL);
+}
+
+
+/*
+ * Perform a rotation to restore balance at the subtree given by depth.
+ *
+ * This routine is used by both insertion and deletion. The return value
+ * indicates:
+ *	 0 : subtree did not change height
+ *	!0 : subtree was reduced in height
+ *
+ * The code is written as if handling left rotations, right rotations are
+ * symmetric and handled by swapping values of variables right/left[_heavy]
+ *
+ * On input balance is the "new" balance at "node". This value is either
+ * -2 or +2.
+ */
+static int
+avl_rotation(avl_tree_t *tree, avl_node_t *node, int balance)
+{
+	int left = !(balance < 0);	/* when balance = -2, left will be 0 */
+	int right = 1 - left;
+	int left_heavy = balance >> 1;
+	int right_heavy = -left_heavy;
+	avl_node_t *parent = AVL_XPARENT(node);
+	avl_node_t *child = node->avl_child[left];
+	avl_node_t *cright;
+	avl_node_t *gchild;
+	avl_node_t *gright;
+	avl_node_t *gleft;
+	int which_child = AVL_XCHILD(node);
+	int child_bal = AVL_XBALANCE(child);
+
+	/* BEGIN CSTYLED */
+	/*
+	 * case 1 : node is overly left heavy, the left child is balanced or
+	 * also left heavy. This requires the following rotation.
+	 *
+	 *                   (node bal:-2)
+	 *                    /           \
+	 *                   /             \
+	 *              (child bal:0 or -1)
+	 *              /    \
+	 *             /      \
+	 *                     cright
+	 *
+	 * becomes:
+	 *
+	 *              (child bal:1 or 0)
+	 *              /        \
+	 *             /          \
+	 *                        (node bal:-1 or 0)
+	 *                         /     \
+	 *                        /       \
+	 *                     cright
+	 *
+	 * we detect this situation by noting that child's balance is not
+	 * right_heavy.
+	 */
+	/* END CSTYLED */
+	if (child_bal != right_heavy) {
+
+		/*
+		 * compute new balance of nodes
+		 *
+		 * If child used to be left heavy (now balanced) we reduced
+		 * the height of this sub-tree -- used in "return...;" below
+		 */
+		child_bal += right_heavy; /* adjust towards right */
+
+		/*
+		 * move "cright" to be node's left child
+		 */
+		cright = child->avl_child[right];
+		node->avl_child[left] = cright;
+		if (cright != NULL) {
+			AVL_SETPARENT(cright, node);
+			AVL_SETCHILD(cright, left);
+		}
+
+		/*
+		 * move node to be child's right child
+		 */
+		child->avl_child[right] = node;
+		AVL_SETBALANCE(node, -child_bal);
+		AVL_SETCHILD(node, right);
+		AVL_SETPARENT(node, child);
+
+		/*
+		 * update the pointer into this subtree
+		 */
+		AVL_SETBALANCE(child, child_bal);
+		AVL_SETCHILD(child, which_child);
+		AVL_SETPARENT(child, parent);
+		if (parent != NULL)
+			parent->avl_child[which_child] = child;
+		else
+			tree->avl_root = child;
+
+		return (child_bal == 0);
+	}
+
+	/* BEGIN CSTYLED */
+	/*
+	 * case 2 : When node is left heavy, but child is right heavy we use
+	 * a different rotation.
+	 *
+	 *                   (node b:-2)
+	 *                    /   \
+	 *                   /     \
+	 *                  /       \
+	 *             (child b:+1)
+	 *              /     \
+	 *             /       \
+	 *                   (gchild b: != 0)
+	 *                     /  \
+	 *                    /    \
+	 *                 gleft   gright
+	 *
+	 * becomes:
+	 *
+	 *              (gchild b:0)
+	 *              /       \
+	 *             /         \
+	 *            /           \
+	 *        (child b:?)   (node b:?)
+	 *         /  \          /   \
+	 *        /    \        /     \
+	 *            gleft   gright
+	 *
+	 * computing the new balances is more complicated. As an example:
+	 *	 if gchild was right_heavy, then child is now left heavy
+	 *		else it is balanced
+	 */
+	/* END CSTYLED */
+	gchild = child->avl_child[right];
+	gleft = gchild->avl_child[left];
+	gright = gchild->avl_child[right];
+
+	/*
+	 * move gright to left child of node and
+	 *
+	 * move gleft to right child of node
+	 */
+	node->avl_child[left] = gright;
+	if (gright != NULL) {
+		AVL_SETPARENT(gright, node);
+		AVL_SETCHILD(gright, left);
+	}
+
+	child->avl_child[right] = gleft;
+	if (gleft != NULL) {
+		AVL_SETPARENT(gleft, child);
+		AVL_SETCHILD(gleft, right);
+	}
+
+	/*
+	 * move child to left child of gchild and
+	 *
+	 * move node to right child of gchild and
+	 *
+	 * fixup parent of all this to point to gchild
+	 */
+	balance = AVL_XBALANCE(gchild);
+	gchild->avl_child[left] = child;
+	AVL_SETBALANCE(child, (balance == right_heavy ? left_heavy : 0));
+	AVL_SETPARENT(child, gchild);
+	AVL_SETCHILD(child, left);
+
+	gchild->avl_child[right] = node;
+	AVL_SETBALANCE(node, (balance == left_heavy ? right_heavy : 0));
+	AVL_SETPARENT(node, gchild);
+	AVL_SETCHILD(node, right);
+
+	AVL_SETBALANCE(gchild, 0);
+	AVL_SETPARENT(gchild, parent);
+	AVL_SETCHILD(gchild, which_child);
+	if (parent != NULL)
+		parent->avl_child[which_child] = gchild;
+	else
+		tree->avl_root = gchild;
+
+	return (1);	/* the new tree is always shorter */
+}
+
+
+/*
+ * Insert a new node into an AVL tree at the specified (from avl_find()) place.
+ *
+ * Newly inserted nodes are always leaf nodes in the tree, since avl_find()
+ * searches out to the leaf positions.  The avl_index_t indicates the node
+ * which will be the parent of the new node.
+ *
+ * After the node is inserted, a single rotation further up the tree may
+ * be necessary to maintain an acceptable AVL balance.
+ */
+void
+avl_insert(avl_tree_t *tree, void *new_data, avl_index_t where)
+{
+	avl_node_t *node;
+	avl_node_t *parent = AVL_INDEX2NODE(where);
+	int old_balance;
+	int new_balance;
+	int which_child = AVL_INDEX2CHILD(where);
+	size_t off = tree->avl_offset;
+
+	ASSERT(tree);
+#ifdef _LP64
+	ASSERT(((uintptr_t)new_data & 0x7) == 0);
+#endif
+
+	node = AVL_DATA2NODE(new_data, off);
+
+	/*
+	 * First, add the node to the tree at the indicated position.
+	 */
+	++tree->avl_numnodes;
+
+	node->avl_child[0] = NULL;
+	node->avl_child[1] = NULL;
+
+	AVL_SETCHILD(node, which_child);
+	AVL_SETBALANCE(node, 0);
+	AVL_SETPARENT(node, parent);
+	if (parent != NULL) {
+		ASSERT(parent->avl_child[which_child] == NULL);
+		parent->avl_child[which_child] = node;
+	} else {
+		ASSERT(tree->avl_root == NULL);
+		tree->avl_root = node;
+	}
+	/*
+	 * Now, back up the tree modifying the balance of all nodes above the
+	 * insertion point. If we get to a highly unbalanced ancestor, we
+	 * need to do a rotation.  If we back out of the tree we are done.
+	 * If we brought any subtree into perfect balance (0), we are also done.
+	 */
+	for (;;) {
+		node = parent;
+		if (node == NULL)
+			return;
+
+		/*
+		 * Compute the new balance
+		 */
+		old_balance = AVL_XBALANCE(node);
+		new_balance = old_balance + avl_child2balance[which_child];
+
+		/*
+		 * If we introduced equal balance, then we are done immediately
+		 */
+		if (new_balance == 0) {
+			AVL_SETBALANCE(node, 0);
+			return;
+		}
+
+		/*
+		 * If both old and new are not zero we went
+		 * from -1 to -2 balance, do a rotation.
+		 */
+		if (old_balance != 0)
+			break;
+
+		AVL_SETBALANCE(node, new_balance);
+		parent = AVL_XPARENT(node);
+		which_child = AVL_XCHILD(node);
+	}
+
+	/*
+	 * perform a rotation to fix the tree and return
+	 */
+	(void) avl_rotation(tree, node, new_balance);
+}
+
+/*
+ * Insert "new_data" in "tree" in the given "direction" either after or
+ * before (AVL_AFTER, AVL_BEFORE) the data "here".
+ *
+ * Insertions can only be done at empty leaf points in the tree, therefore
+ * if the given child of the node is already present we move to either
+ * the AVL_PREV or AVL_NEXT and reverse the insertion direction. Since
+ * every other node in the tree is a leaf, this always works.
+ *
+ * To help developers using this interface, we assert that the new node
+ * is correctly ordered at every step of the way in DEBUG kernels.
+ */
+void
+avl_insert_here(
+	avl_tree_t *tree,
+	void *new_data,
+	void *here,
+	int direction)
+{
+	avl_node_t *node;
+	int child = direction;	/* rely on AVL_BEFORE == 0, AVL_AFTER == 1 */
+#ifdef DEBUG
+	int diff;
+#endif
+
+	ASSERT(tree != NULL);
+	ASSERT(new_data != NULL);
+	ASSERT(here != NULL);
+	ASSERT(direction == AVL_BEFORE || direction == AVL_AFTER);
+
+	/*
+	 * If corresponding child of node is not NULL, go to the neighboring
+	 * node and reverse the insertion direction.
+	 */
+	node = AVL_DATA2NODE(here, tree->avl_offset);
+
+#ifdef DEBUG
+	diff = tree->avl_compar(new_data, here);
+	ASSERT(-1 <= diff && diff <= 1);
+	ASSERT(diff != 0);
+	ASSERT(diff > 0 ? child == 1 : child == 0);
+#endif
+
+	if (node->avl_child[child] != NULL) {
+		node = node->avl_child[child];
+		child = 1 - child;
+		while (node->avl_child[child] != NULL) {
+#ifdef DEBUG
+			diff = tree->avl_compar(new_data,
+			    AVL_NODE2DATA(node, tree->avl_offset));
+			ASSERT(-1 <= diff && diff <= 1);
+			ASSERT(diff != 0);
+			ASSERT(diff > 0 ? child == 1 : child == 0);
+#endif
+			node = node->avl_child[child];
+		}
+#ifdef DEBUG
+		diff = tree->avl_compar(new_data,
+		    AVL_NODE2DATA(node, tree->avl_offset));
+		ASSERT(-1 <= diff && diff <= 1);
+		ASSERT(diff != 0);
+		ASSERT(diff > 0 ? child == 1 : child == 0);
+#endif
+	}
+	ASSERT(node->avl_child[child] == NULL);
+
+	avl_insert(tree, new_data, AVL_MKINDEX(node, child));
+}
+
+/*
+ * Add a new node to an AVL tree.
+ */
+void
+avl_add(avl_tree_t *tree, void *new_node)
+{
+	avl_index_t where;
+
+	/*
+	 * This is unfortunate.  We want to call panic() here, even for
+	 * non-DEBUG kernels.  In userland, however, we can't depend on anything
+	 * in libc or else the rtld build process gets confused.  So, all we can
+	 * do in userland is resort to a normal ASSERT().
+	 */
+	if (avl_find(tree, new_node, &where) != NULL)
+#ifdef _KERNEL
+		panic("avl_find() succeeded inside avl_add()");
+#else
+		ASSERT(0);
+#endif
+	avl_insert(tree, new_node, where);
+}
+
+/*
+ * Delete a node from the AVL tree.  Deletion is similar to insertion, but
+ * with 2 complications.
+ *
+ * First, we may be deleting an interior node. Consider the following subtree:
+ *
+ *     d           c            c
+ *    / \         / \          / \
+ *   b   e       b   e        b   e
+ *  / \	        / \          /
+ * a   c       a            a
+ *
+ * When we are deleting node (d), we find and bring up an adjacent valued leaf
+ * node, say (c), to take the interior node's place. In the code this is
+ * handled by temporarily swapping (d) and (c) in the tree and then using
+ * common code to delete (d) from the leaf position.
+ *
+ * Secondly, an interior deletion from a deep tree may require more than one
+ * rotation to fix the balance. This is handled by moving up the tree through
+ * parents and applying rotations as needed. The return value from
+ * avl_rotation() is used to detect when a subtree did not change overall
+ * height due to a rotation.
+ */
+void
+avl_remove(avl_tree_t *tree, void *data)
+{
+	avl_node_t *delete;
+	avl_node_t *parent;
+	avl_node_t *node;
+	avl_node_t tmp;
+	int old_balance;
+	int new_balance;
+	int left;
+	int right;
+	int which_child;
+	size_t off = tree->avl_offset;
+
+	ASSERT(tree);
+
+	delete = AVL_DATA2NODE(data, off);
+
+	/*
+	 * Deletion is easiest with a node that has at most 1 child.
+	 * We swap a node with 2 children with a sequentially valued
+	 * neighbor node. That node will have at most 1 child. Note this
+	 * has no effect on the ordering of the remaining nodes.
+	 *
+	 * As an optimization, we choose the greater neighbor if the tree
+	 * is right heavy, otherwise the left neighbor. This reduces the
+	 * number of rotations needed.
+	 */
+	if (delete->avl_child[0] != NULL && delete->avl_child[1] != NULL) {
+
+		/*
+		 * choose node to swap from whichever side is taller
+		 */
+		old_balance = AVL_XBALANCE(delete);
+		left = avl_balance2child[old_balance + 1];
+		right = 1 - left;
+
+		/*
+		 * get to the previous value'd node
+		 * (down 1 left, as far as possible right)
+		 */
+		for (node = delete->avl_child[left];
+		    node->avl_child[right] != NULL;
+		    node = node->avl_child[right])
+			;
+
+		/*
+		 * create a temp placeholder for 'node'
+		 * move 'node' to delete's spot in the tree
+		 */
+		tmp = *node;
+
+		*node = *delete;
+		if (node->avl_child[left] == node)
+			node->avl_child[left] = &tmp;
+
+		parent = AVL_XPARENT(node);
+		if (parent != NULL)
+			parent->avl_child[AVL_XCHILD(node)] = node;
+		else
+			tree->avl_root = node;
+		AVL_SETPARENT(node->avl_child[left], node);
+		AVL_SETPARENT(node->avl_child[right], node);
+
+		/*
+		 * Put tmp where node used to be (just temporary).
+		 * It always has a parent and at most 1 child.
+		 */
+		delete = &tmp;
+		parent = AVL_XPARENT(delete);
+		parent->avl_child[AVL_XCHILD(delete)] = delete;
+		which_child = (delete->avl_child[1] != 0);
+		if (delete->avl_child[which_child] != NULL)
+			AVL_SETPARENT(delete->avl_child[which_child], delete);
+	}
+
+
+	/*
+	 * Here we know "delete" is at least partially a leaf node. It can
+	 * be easily removed from the tree.
+	 */
+	ASSERT(tree->avl_numnodes > 0);
+	--tree->avl_numnodes;
+	parent = AVL_XPARENT(delete);
+	which_child = AVL_XCHILD(delete);
+	if (delete->avl_child[0] != NULL)
+		node = delete->avl_child[0];
+	else
+		node = delete->avl_child[1];
+
+	/*
+	 * Connect parent directly to node (leaving out delete).
+	 */
+	if (node != NULL) {
+		AVL_SETPARENT(node, parent);
+		AVL_SETCHILD(node, which_child);
+	}
+	if (parent == NULL) {
+		tree->avl_root = node;
+		return;
+	}
+	parent->avl_child[which_child] = node;
+
+
+	/*
+	 * Since the subtree is now shorter, begin adjusting parent balances
+	 * and performing any needed rotations.
+	 */
+	do {
+
+		/*
+		 * Move up the tree and adjust the balance
+		 *
+		 * Capture the parent and which_child values for the next
+		 * iteration before any rotations occur.
+		 */
+		node = parent;
+		old_balance = AVL_XBALANCE(node);
+		new_balance = old_balance - avl_child2balance[which_child];
+		parent = AVL_XPARENT(node);
+		which_child = AVL_XCHILD(node);
+
+		/*
+		 * If a node was in perfect balance but isn't anymore then
+		 * we can stop, since the height didn't change above this point
+		 * due to a deletion.
+		 */
+		if (old_balance == 0) {
+			AVL_SETBALANCE(node, new_balance);
+			break;
+		}
+
+		/*
+		 * If the new balance is zero, we don't need to rotate
+		 * else
+		 * need a rotation to fix the balance.
+		 * If the rotation doesn't change the height
+		 * of the sub-tree we have finished adjusting.
+		 */
+		if (new_balance == 0)
+			AVL_SETBALANCE(node, new_balance);
+		else if (!avl_rotation(tree, node, new_balance))
+			break;
+	} while (parent != NULL);
+}
+
+/*
+ * initialize a new AVL tree
+ */
+void
+avl_create(avl_tree_t *tree, int (*compar) (const void *, const void *),
+    size_t size, size_t offset)
+{
+	ASSERT(tree);
+	ASSERT(compar);
+	ASSERT(size > 0);
+	ASSERT(size >= offset + sizeof (avl_node_t));
+#ifdef _LP64
+	ASSERT((offset & 0x7) == 0);
+#endif
+
+	tree->avl_compar = compar;
+	tree->avl_root = NULL;
+	tree->avl_numnodes = 0;
+	tree->avl_size = size;
+	tree->avl_offset = offset;
+}
+
+/*
+ * Delete a tree.
+ */
+/* ARGSUSED */
+void
+avl_destroy(avl_tree_t *tree)
+{
+	ASSERT(tree);
+	ASSERT(tree->avl_numnodes == 0);
+	ASSERT(tree->avl_root == NULL);
+}
+
+
+/*
+ * Return the number of nodes in an AVL tree.
+ */
+ulong_t
+avl_numnodes(avl_tree_t *tree)
+{
+	ASSERT(tree);
+	return (tree->avl_numnodes);
+}
+
+
+#define	CHILDBIT	(1L)
+
+/*
+ * Post-order tree walk used to visit all tree nodes and destroy the tree
+ * in post order. This is used for destroying a tree w/o paying any cost
+ * for rebalancing it.
+ *
+ * example:
+ *
+ *	void *cookie = NULL;
+ *	my_data_t *node;
+ *
+ *	while ((node = avl_destroy_nodes(tree, &cookie)) != NULL)
+ *		free(node);
+ *	avl_destroy(tree);
+ *
+ * The cookie is really an avl_node_t to the current node's parent and
+ * an indication of which child you looked at last.
+ *
+ * On input, a cookie value of CHILDBIT indicates the tree is done.
+ */
+void *
+avl_destroy_nodes(avl_tree_t *tree, void **cookie)
+{
+	avl_node_t	*node;
+	avl_node_t	*parent;
+	int		child;
+	void		*first;
+	size_t		off = tree->avl_offset;
+
+	/*
+	 * Initial calls go to the first node or it's right descendant.
+	 */
+	if (*cookie == NULL) {
+		first = avl_first(tree);
+
+		/*
+		 * deal with an empty tree
+		 */
+		if (first == NULL) {
+			*cookie = (void *)CHILDBIT;
+			return (NULL);
+		}
+
+		node = AVL_DATA2NODE(first, off);
+		parent = AVL_XPARENT(node);
+		goto check_right_side;
+	}
+
+	/*
+	 * If there is no parent to return to we are done.
+	 */
+	parent = (avl_node_t *)((uintptr_t)(*cookie) & ~CHILDBIT);
+	if (parent == NULL) {
+		if (tree->avl_root != NULL) {
+			ASSERT(tree->avl_numnodes == 1);
+			tree->avl_root = NULL;
+			tree->avl_numnodes = 0;
+		}
+		return (NULL);
+	}
+
+	/*
+	 * Remove the child pointer we just visited from the parent and tree.
+	 */
+	child = (uintptr_t)(*cookie) & CHILDBIT;
+	parent->avl_child[child] = NULL;
+	ASSERT(tree->avl_numnodes > 1);
+	--tree->avl_numnodes;
+
+	/*
+	 * If we just did a right child or there isn't one, go up to parent.
+	 */
+	if (child == 1 || parent->avl_child[1] == NULL) {
+		node = parent;
+		parent = AVL_XPARENT(parent);
+		goto done;
+	}
+
+	/*
+	 * Do parent's right child, then leftmost descendent.
+	 */
+	node = parent->avl_child[1];
+	while (node->avl_child[0] != NULL) {
+		parent = node;
+		node = node->avl_child[0];
+	}
+
+	/*
+	 * If here, we moved to a left child. It may have one
+	 * child on the right (when balance == +1).
+	 */
+check_right_side:
+	if (node->avl_child[1] != NULL) {
+		ASSERT(AVL_XBALANCE(node) == 1);
+		parent = node;
+		node = node->avl_child[1];
+		ASSERT(node->avl_child[0] == NULL &&
+		    node->avl_child[1] == NULL);
+	} else {
+		ASSERT(AVL_XBALANCE(node) <= 0);
+	}
+
+done:
+	if (parent == NULL) {
+		*cookie = (void *)CHILDBIT;
+		ASSERT(node == tree->avl_root);
+	} else {
+		*cookie = (void *)((uintptr_t)parent | AVL_XCHILD(node));
+	}
+
+	return (AVL_NODE2DATA(node, off));
+}
--- a/zfs/lib/libavl/include/Makefile.in
+++ b/zfs/lib/libavl/include/Makefile.in
@ -0,0 +1 @@
+subdir-m += sys
--- a/zfs/lib/libavl/include/sys/Makefile.in
+++ b/zfs/lib/libavl/include/sys/Makefile.in
@ -0,0 +1 @@
+DISTFILES = avl.h avl_impl.h
--- a/zfs/lib/libavl/include/sys/avl.h
+++ b/zfs/lib/libavl/include/sys/avl.h
@ -0,0 +1,298 @@
+/*
+ * CDDL HEADER START
+ *
+ * The contents of this file are subject to the terms of the
+ * Common Development and Distribution License, Version 1.0 only
+ * (the "License").  You may not use this file except in compliance
+ * with the License.
+ *
+ * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
+ * or http://www.opensolaris.org/os/licensing.
+ * See the License for the specific language governing permissions
+ * and limitations under the License.
+ *
+ * When distributing Covered Code, include this CDDL HEADER in each
+ * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
+ * If applicable, add the following below this CDDL HEADER, with the
+ * fields enclosed by brackets "[]" replaced with your own identifying
+ * information: Portions Copyright [yyyy] [name of copyright owner]
+ *
+ * CDDL HEADER END
+ */
+/*
+ * Copyright 2005 Sun Microsystems, Inc.  All rights reserved.
+ * Use is subject to license terms.
+ */
+
+#ifndef	_AVL_H
+#define	_AVL_H
+
+
+
+/*
+ * This is a private header file.  Applications should not directly include
+ * this file.
+ */
+
+#ifdef	__cplusplus
+extern "C" {
+#endif
+
+#include <sys/avl_impl.h>
+
+/*
+ * This is a generic implemenatation of AVL trees for use in the Solaris kernel.
+ * The interfaces provide an efficient way of implementing an ordered set of
+ * data structures.
+ *
+ * AVL trees provide an alternative to using an ordered linked list. Using AVL
+ * trees will usually be faster, however they requires more storage. An ordered
+ * linked list in general requires 2 pointers in each data structure. The
+ * AVL tree implementation uses 3 pointers. The following chart gives the
+ * approximate performance of operations with the different approaches:
+ *
+ *	Operation	 Link List	AVL tree
+ *	---------	 --------	--------
+ *	lookup		   O(n)		O(log(n))
+ *
+ *	insert 1 node	 constant	constant
+ *
+ *	delete 1 node	 constant	between constant and O(log(n))
+ *
+ *	delete all nodes   O(n)		O(n)
+ *
+ *	visit the next
+ *	or prev node	 constant	between constant and O(log(n))
+ *
+ *
+ * The data structure nodes are anchored at an "avl_tree_t" (the equivalent
+ * of a list header) and the individual nodes will have a field of
+ * type "avl_node_t" (corresponding to list pointers).
+ *
+ * The type "avl_index_t" is used to indicate a position in the list for
+ * certain calls.
+ *
+ * The usage scenario is generally:
+ *
+ * 1. Create the list/tree with: avl_create()
+ *
+ * followed by any mixture of:
+ *
+ * 2a. Insert nodes with: avl_add(), or avl_find() and avl_insert()
+ *
+ * 2b. Visited elements with:
+ *	 avl_first() - returns the lowest valued node
+ *	 avl_last() - returns the highest valued node
+ *	 AVL_NEXT() - given a node go to next higher one
+ *	 AVL_PREV() - given a node go to previous lower one
+ *
+ * 2c.  Find the node with the closest value either less than or greater
+ *	than a given value with avl_nearest().
+ *
+ * 2d. Remove individual nodes from the list/tree with avl_remove().
+ *
+ * and finally when the list is being destroyed
+ *
+ * 3. Use avl_destroy_nodes() to quickly process/free up any remaining nodes.
+ *    Note that once you use avl_destroy_nodes(), you can no longer
+ *    use any routine except avl_destroy_nodes() and avl_destoy().
+ *
+ * 4. Use avl_destroy() to destroy the AVL tree itself.
+ *
+ * Any locking for multiple thread access is up to the user to provide, just
+ * as is needed for any linked list implementation.
+ */
+
+
+/*
+ * Type used for the root of the AVL tree.
+ */
+typedef struct avl_tree avl_tree_t;
+
+/*
+ * The data nodes in the AVL tree must have a field of this type.
+ */
+typedef struct avl_node avl_node_t;
+
+/*
+ * An opaque type used to locate a position in the tree where a node
+ * would be inserted.
+ */
+typedef uintptr_t avl_index_t;
+
+
+/*
+ * Direction constants used for avl_nearest().
+ */
+#define	AVL_BEFORE	(0)
+#define	AVL_AFTER	(1)
+
+
+
+/*
+ * Prototypes
+ *
+ * Where not otherwise mentioned, "void *" arguments are a pointer to the
+ * user data structure which must contain a field of type avl_node_t.
+ *
+ * Also assume the user data structures looks like:
+ *	stuct my_type {
+ *		...
+ *		avl_node_t	my_link;
+ *		...
+ *	};
+ */
+
+/*
+ * Initialize an AVL tree. Arguments are:
+ *
+ * tree   - the tree to be initialized
+ * compar - function to compare two nodes, it must return exactly: -1, 0, or +1
+ *          -1 for <, 0 for ==, and +1 for >
+ * size   - the value of sizeof(struct my_type)
+ * offset - the value of OFFSETOF(struct my_type, my_link)
+ */
+extern void avl_create(avl_tree_t *tree,
+	int (*compar) (const void *, const void *), size_t size, size_t offset);
+
+
+/*
+ * Find a node with a matching value in the tree. Returns the matching node
+ * found. If not found, it returns NULL and then if "where" is not NULL it sets
+ * "where" for use with avl_insert() or avl_nearest().
+ *
+ * node   - node that has the value being looked for
+ * where  - position for use with avl_nearest() or avl_insert(), may be NULL
+ */
+extern void *avl_find(avl_tree_t *tree, void *node, avl_index_t *where);
+
+/*
+ * Insert a node into the tree.
+ *
+ * node   - the node to insert
+ * where  - position as returned from avl_find()
+ */
+extern void avl_insert(avl_tree_t *tree, void *node, avl_index_t where);
+
+/*
+ * Insert "new_data" in "tree" in the given "direction" either after
+ * or before the data "here".
+ *
+ * This might be usefull for avl clients caching recently accessed
+ * data to avoid doing avl_find() again for insertion.
+ *
+ * new_data	- new data to insert
+ * here  	- existing node in "tree"
+ * direction	- either AVL_AFTER or AVL_BEFORE the data "here".
+ */
+extern void avl_insert_here(avl_tree_t *tree, void *new_data, void *here,
+    int direction);
+
+
+/*
+ * Return the first or last valued node in the tree. Will return NULL
+ * if the tree is empty.
+ *
+ */
+extern void *avl_first(avl_tree_t *tree);
+extern void *avl_last(avl_tree_t *tree);
+
+
+/*
+ * Return the next or previous valued node in the tree.
+ * AVL_NEXT() will return NULL if at the last node.
+ * AVL_PREV() will return NULL if at the first node.
+ *
+ * node   - the node from which the next or previous node is found
+ */
+#define	AVL_NEXT(tree, node)	avl_walk(tree, node, AVL_AFTER)
+#define	AVL_PREV(tree, node)	avl_walk(tree, node, AVL_BEFORE)
+
+
+/*
+ * Find the node with the nearest value either greater or less than
+ * the value from a previous avl_find(). Returns the node or NULL if
+ * there isn't a matching one.
+ *
+ * where     - position as returned from avl_find()
+ * direction - either AVL_BEFORE or AVL_AFTER
+ *
+ * EXAMPLE get the greatest node that is less than a given value:
+ *
+ *	avl_tree_t *tree;
+ *	struct my_data look_for_value = {....};
+ *	struct my_data *node;
+ *	struct my_data *less;
+ *	avl_index_t where;
+ *
+ *	node = avl_find(tree, &look_for_value, &where);
+ *	if (node != NULL)
+ *		less = AVL_PREV(tree, node);
+ *	else
+ *		less = avl_nearest(tree, where, AVL_BEFORE);
+ */
+extern void *avl_nearest(avl_tree_t *tree, avl_index_t where, int direction);
+
+
+/*
+ * Add a single node to the tree.
+ * The node must not be in the tree, and it must not
+ * compare equal to any other node already in the tree.
+ *
+ * node   - the node to add
+ */
+extern void avl_add(avl_tree_t *tree, void *node);
+
+
+/*
+ * Remove a single node from the tree.  The node must be in the tree.
+ *
+ * node   - the node to remove
+ */
+extern void avl_remove(avl_tree_t *tree, void *node);
+
+
+/*
+ * Return the number of nodes in the tree
+ */
+extern ulong_t avl_numnodes(avl_tree_t *tree);
+
+
+/*
+ * Used to destroy any remaining nodes in a tree. The cookie argument should
+ * be initialized to NULL before the first call. Returns a node that has been
+ * removed from the tree and may be free()'d. Returns NULL when the tree is
+ * empty.
+ *
+ * Once you call avl_destroy_nodes(), you can only continuing calling it and
+ * finally avl_destroy(). No other AVL routines will be valid.
+ *
+ * cookie - a "void *" used to save state between calls to avl_destroy_nodes()
+ *
+ * EXAMPLE:
+ *	avl_tree_t *tree;
+ *	struct my_data *node;
+ *	void *cookie;
+ *
+ *	cookie = NULL;
+ *	while ((node = avl_destroy_nodes(tree, &cookie)) != NULL)
+ *		free(node);
+ *	avl_destroy(tree);
+ */
+extern void *avl_destroy_nodes(avl_tree_t *tree, void **cookie);
+
+
+/*
+ * Final destroy of an AVL tree. Arguments are:
+ *
+ * tree   - the empty tree to destroy
+ */
+extern void avl_destroy(avl_tree_t *tree);
+
+
+
+#ifdef	__cplusplus
+}
+#endif
+
+#endif	/* _AVL_H */
--- a/zfs/lib/libavl/include/sys/avl_impl.h
+++ b/zfs/lib/libavl/include/sys/avl_impl.h
@ -0,0 +1,164 @@
+/*
+ * CDDL HEADER START
+ *
+ * The contents of this file are subject to the terms of the
+ * Common Development and Distribution License, Version 1.0 only
+ * (the "License").  You may not use this file except in compliance
+ * with the License.
+ *
+ * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
+ * or http://www.opensolaris.org/os/licensing.
+ * See the License for the specific language governing permissions
+ * and limitations under the License.
+ *
+ * When distributing Covered Code, include this CDDL HEADER in each
+ * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
+ * If applicable, add the following below this CDDL HEADER, with the
+ * fields enclosed by brackets "[]" replaced with your own identifying
+ * information: Portions Copyright [yyyy] [name of copyright owner]
+ *
+ * CDDL HEADER END
+ */
+/*
+ * Copyright 2004 Sun Microsystems, Inc.  All rights reserved.
+ * Use is subject to license terms.
+ */
+
+#ifndef	_AVL_IMPL_H
+#define	_AVL_IMPL_H
+
+
+
+/*
+ * This is a private header file.  Applications should not directly include
+ * this file.
+ */
+
+#include <sys/types.h>
+
+#ifdef	__cplusplus
+extern "C" {
+#endif
+
+
+/*
+ * generic AVL tree implementation for kernel use
+ *
+ * There are 5 pieces of information stored for each node in an AVL tree
+ *
+ * 	pointer to less than child
+ * 	pointer to greater than child
+ * 	a pointer to the parent of this node
+ *	an indication  [0/1]  of which child I am of my parent
+ * 	a "balance" (-1, 0, +1)  indicating which child tree is taller
+ *
+ * Since they only need 3 bits, the last two fields are packed into the
+ * bottom bits of the parent pointer on 64 bit machines to save on space.
+ */
+
+#ifndef _LP64
+
+struct avl_node {
+	struct avl_node *avl_child[2];	/* left/right children */
+	struct avl_node *avl_parent;	/* this node's parent */
+	unsigned short avl_child_index;	/* my index in parent's avl_child[] */
+	short avl_balance;		/* balance value: -1, 0, +1 */
+};
+
+#define	AVL_XPARENT(n)		((n)->avl_parent)
+#define	AVL_SETPARENT(n, p)	((n)->avl_parent = (p))
+
+#define	AVL_XCHILD(n)		((n)->avl_child_index)
+#define	AVL_SETCHILD(n, c)	((n)->avl_child_index = (unsigned short)(c))
+
+#define	AVL_XBALANCE(n)		((n)->avl_balance)
+#define	AVL_SETBALANCE(n, b)	((n)->avl_balance = (short)(b))
+
+#else /* _LP64 */
+
+/*
+ * for 64 bit machines, avl_pcb contains parent pointer, balance and child_index
+ * values packed in the following manner:
+ *
+ * |63                                  3|        2        |1          0 |
+ * |-------------------------------------|-----------------|-------------|
+ * |      avl_parent hi order bits       | avl_child_index | avl_balance |
+ * |                                     |                 |     + 1     |
+ * |-------------------------------------|-----------------|-------------|
+ *
+ */
+struct avl_node {
+	struct avl_node *avl_child[2];	/* left/right children nodes */
+	uintptr_t avl_pcb;		/* parent, child_index, balance */
+};
+
+/*
+ * macros to extract/set fields in avl_pcb
+ *
+ * pointer to the parent of the current node is the high order bits
+ */
+#define	AVL_XPARENT(n)		((struct avl_node *)((n)->avl_pcb & ~7))
+#define	AVL_SETPARENT(n, p)						\
+	((n)->avl_pcb = (((n)->avl_pcb & 7) | (uintptr_t)(p)))
+
+/*
+ * index of this node in its parent's avl_child[]: bit #2
+ */
+#define	AVL_XCHILD(n)		(((n)->avl_pcb >> 2) & 1)
+#define	AVL_SETCHILD(n, c)						\
+	((n)->avl_pcb = (uintptr_t)(((n)->avl_pcb & ~4) | ((c) << 2)))
+
+/*
+ * balance indication for a node, lowest 2 bits. A valid balance is
+ * -1, 0, or +1, and is encoded by adding 1 to the value to get the
+ * unsigned values of 0, 1, 2.
+ */
+#define	AVL_XBALANCE(n)		((int)(((n)->avl_pcb & 3) - 1))
+#define	AVL_SETBALANCE(n, b)						\
+	((n)->avl_pcb = (uintptr_t)((((n)->avl_pcb & ~3) | ((b) + 1))))
+
+#endif /* _LP64 */
+
+
+
+/*
+ * switch between a node and data pointer for a given tree
+ * the value of "o" is tree->avl_offset
+ */
+#define	AVL_NODE2DATA(n, o)	((void *)((uintptr_t)(n) - (o)))
+#define	AVL_DATA2NODE(d, o)	((struct avl_node *)((uintptr_t)(d) + (o)))
+
+
+
+/*
+ * macros used to create/access an avl_index_t
+ */
+#define	AVL_INDEX2NODE(x)	((avl_node_t *)((x) & ~1))
+#define	AVL_INDEX2CHILD(x)	((x) & 1)
+#define	AVL_MKINDEX(n, c)	((avl_index_t)(n) | (c))
+
+
+/*
+ * The tree structure. The fields avl_root, avl_compar, and avl_offset come
+ * first since they are needed for avl_find().  We want them to fit into
+ * a single 64 byte cache line to make avl_find() as fast as possible.
+ */
+struct avl_tree {
+	struct avl_node *avl_root;	/* root node in tree */
+	int (*avl_compar)(const void *, const void *);
+	size_t avl_offset;		/* offsetof(type, avl_link_t field) */
+	ulong_t avl_numnodes;		/* number of nodes in the tree */
+	size_t avl_size;		/* sizeof user type struct */
+};
+
+
+/*
+ * This will only by used via AVL_NEXT() or AVL_PREV()
+ */
+extern void *avl_walk(struct avl_tree *, void *, int);
+
+#ifdef	__cplusplus
+}
+#endif
+
+#endif	/* _AVL_IMPL_H */
--- a/zfs/lib/libdmu-ctl/Makefile.in
+++ b/zfs/lib/libdmu-ctl/Makefile.in
@ -0,0 +1,28 @@
+# NOTE: dctl_client.c, dctl_common.c, dctl_server.c, dctl_thrpool.c unused
+# by kernel port.  Potentially they should just be removed if we don't care
+# able user space lustre intergration from this source base.
+
+# NOTE: For clarity this directly should simply be renamed libzpl and
+# the full kernel implementation should be minimally stubbed out.
+
+subdir-m += include
+DISTFILES  = dctl_client.c dctl_common.c dctl_server.c dctl_thrpool.c
+DISTFILES += dmu_send.c rrwlock.c zfs_acl.c zfs_ctldir.c
+DISTFILES += zfs_dir.c zfs_fuid.c zfs_ioctl.c zfs_log.c zfs_replay.c
+DISTFILES += zfs_rlock.c zfs_vfsops.c zfs_vnops.c zvol.c
+
+MODULE := zctl
+
+EXTRA_CFLAGS  = @KERNELCPPFLAGS@
+EXTRA_CFLAGS += -I@LIBDIR@/libzcommon/include
+EXTRA_CFLAGS += -I@LIBDIR@/libdmu-ctl/include
+EXTRA_CFLAGS += -I@LIBDIR@/libavl/include
+EXTRA_CFLAGS += -I@LIBDIR@/libport/include
+EXTRA_CFLAGS += -I@LIBDIR@/libnvpair/include
+
+obj-m := ${MODULE}.o
+
+${MODULE}-objs += zvol.o	# Volume emulation interface
+${MODULE}-objs += zfs_ioctl.o	# /dev/zfs_ioctl interface
+${MODULE}-objs += zfs_vfsops.o
+${MODULE}-objs += dmu_send.o
--- a/zfs/lib/libdmu-ctl/dctl_client.c
+++ b/zfs/lib/libdmu-ctl/dctl_client.c
@ -0,0 +1,263 @@
+/*
+ * CDDL HEADER START
+ *
+ * The contents of this file are subject to the terms of the
+ * Common Development and Distribution License, Version 1.0 only
+ * (the "License").  You may not use this file except in compliance
+ * with the License.
+ *
+ * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
+ * or http://www.opensolaris.org/os/licensing.
+ * See the License for the specific language governing permissions
+ * and limitations under the License.
+ *
+ * When distributing Covered Code, include this CDDL HEADER in each
+ * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
+ * If applicable, add the following below this CDDL HEADER, with the
+ * fields enclosed by brackets "[]" replaced with your own identifying
+ * information: Portions Copyright [yyyy] [name of copyright owner]
+ *
+ * CDDL HEADER END
+ */
+/*
+ * Copyright 2008 Sun Microsystems, Inc.  All rights reserved.
+ * Use is subject to license terms.
+ */
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <ftw.h>
+#include <errno.h>
+#include <unistd.h>
+#include <sys/types.h>
+#include <sys/socket.h>
+#include <sys/un.h>
+#include <sys/debug.h>
+
+#include <sys/dmu_ctl.h>
+#include <sys/dmu_ctl_impl.h>
+
+/*
+ * Try to connect to the socket given in path.
+ *
+ * For nftw() convenience, returns 0 if unsuccessful, otherwise
+ * returns the socket descriptor.
+ */
+static int try_connect(const char *path)
+{
+	struct sockaddr_un name;
+	int sock;
+
+	sock = socket(PF_UNIX, SOCK_STREAM, 0);
+	if (sock == -1) {
+		perror("socket");
+		return 0;
+	}
+
+	/*
+	 * The socket fd cannot be 0 otherwise nftw() will not interpret the
+	 * return code correctly.
+	 */
+	VERIFY(sock != 0);
+
+	name.sun_family = AF_UNIX;
+	strncpy(name.sun_path, path, sizeof(name.sun_path));
+
+	name.sun_path[sizeof(name.sun_path) - 1] = '\0';
+
+	if (connect(sock, (struct sockaddr *) &name, sizeof(name)) == -1) {
+		close(sock);
+		return 0;
+	}
+
+	return sock;
+}
+
+/*
+ * nftw() callback.
+ */
+static int nftw_cb(const char *fpath, const struct stat *sb, int typeflag,
+    struct FTW *ftwbuf)
+{
+	if (!S_ISSOCK(sb->st_mode))
+		return 0;
+
+	if (strcmp(&fpath[ftwbuf->base], SOCKNAME) != 0)
+		return 0;
+
+	return try_connect(fpath);
+}
+
+/*
+ * For convenience, if check_subdirs is true we walk the directory tree to
+ * find a good socket.
+ */
+int dctlc_connect(const char *dir, boolean_t check_subdirs)
+{
+	char *fpath;
+	int fd;
+
+	if (check_subdirs)
+		fd = nftw(dir, nftw_cb, 10, FTW_PHYS);
+	else {
+		fpath = malloc(strlen(dir) + strlen(SOCKNAME) + 2);
+		if (fpath == NULL)
+			return -1;
+
+		strcpy(fpath, dir);
+		strcat(fpath, "/" SOCKNAME);
+
+		fd = try_connect(fpath);
+
+		free(fpath);
+	}
+
+	return fd == 0 ? -1 : fd;
+}
+
+void dctlc_disconnect(int fd)
+{
+	(void) shutdown(fd, SHUT_RDWR);
+}
+
+static int dctl_reply_copyin(int fd, dctl_cmd_t *cmd)
+{
+	return dctl_send_data(fd, (void *)(uintptr_t) cmd->u.dcmd_copy.ptr,
+	    cmd->u.dcmd_copy.size);
+}
+
+static int dctl_reply_copyinstr(int fd, dctl_cmd_t *cmd)
+{
+	dctl_cmd_t reply;
+	char *from;
+	size_t len, buflen, to_copy;
+	int error;
+
+	reply.dcmd_msg = DCTL_GEN_REPLY;
+
+	from = (char *)(uintptr_t) cmd->u.dcmd_copy.ptr;
+
+	buflen = cmd->u.dcmd_copy.size;
+	to_copy = strnlen(from, buflen - 1);
+
+	reply.u.dcmd_reply.rc = from[to_copy] == '\0' ? 0 : ENAMETOOLONG;
+	reply.u.dcmd_reply.size = to_copy;
+
+	error = dctl_send_msg(fd, &reply);
+
+	if (!error && to_copy > 0)
+		error = dctl_send_data(fd, from, to_copy);
+
+	return error;
+}
+
+static int dctl_reply_copyout(int fd, dctl_cmd_t *cmd)
+{
+	return dctl_read_data(fd, (void *)(uintptr_t) cmd->u.dcmd_copy.ptr,
+	    cmd->u.dcmd_copy.size);
+}
+
+static int dctl_reply_fd_read(int fd, dctl_cmd_t *cmd)
+{
+	dctl_cmd_t reply;
+	void *buf;
+	int error;
+	ssize_t rrc, size = cmd->u.dcmd_fd_io.size;
+
+	buf = malloc(size);
+	if (buf == NULL)
+		return ENOMEM;
+
+	rrc = read(cmd->u.dcmd_fd_io.fd, buf, size);
+
+	reply.dcmd_msg = DCTL_GEN_REPLY;
+	reply.u.dcmd_reply.rc = rrc == -1 ? errno : 0;
+	reply.u.dcmd_reply.size = rrc;
+
+	error = dctl_send_msg(fd, &reply);
+
+	if (!error && rrc > 0)
+		error = dctl_send_data(fd, buf, rrc);
+
+out:
+	free(buf);
+
+	return error;
+}
+
+static int dctl_reply_fd_write(int fd, dctl_cmd_t *cmd)
+{
+	dctl_cmd_t reply;
+	void *buf;
+	int error;
+	ssize_t wrc, size = cmd->u.dcmd_fd_io.size;
+
+	buf = malloc(size);
+	if (buf == NULL)
+		return ENOMEM;
+
+	error = dctl_read_data(fd, buf, size);
+	if (error)
+		goto out;
+
+	wrc = write(cmd->u.dcmd_fd_io.fd, buf, size);
+
+	reply.dcmd_msg = DCTL_GEN_REPLY;
+	reply.u.dcmd_reply.rc = wrc == -1 ? errno : 0;
+	reply.u.dcmd_reply.size = wrc;
+
+	error = dctl_send_msg(fd, &reply);
+
+out:
+	free(buf);
+
+	return error;
+}
+
+int dctlc_ioctl(int fd, int32_t request, void *arg)
+{
+	int error;
+	dctl_cmd_t cmd;
+
+	ASSERT(fd != 0);
+
+	cmd.dcmd_msg = DCTL_IOCTL;
+
+	cmd.u.dcmd_ioctl.cmd = request;
+	cmd.u.dcmd_ioctl.arg = (uintptr_t) arg;
+
+	error = dctl_send_msg(fd, &cmd);
+
+	while (!error && (error = dctl_read_msg(fd, &cmd)) == 0) {
+		switch (cmd.dcmd_msg) {
+			case DCTL_IOCTL_REPLY:
+				error = cmd.u.dcmd_reply.rc;
+				goto out;
+			case DCTL_COPYIN:
+				error = dctl_reply_copyin(fd, &cmd);
+				break;
+			case DCTL_COPYINSTR:
+				error = dctl_reply_copyinstr(fd, &cmd);
+				break;
+			case DCTL_COPYOUT:
+				error = dctl_reply_copyout(fd, &cmd);
+				break;
+			case DCTL_FD_READ:
+				error = dctl_reply_fd_read(fd, &cmd);
+				break;
+			case DCTL_FD_WRITE:
+				error = dctl_reply_fd_write(fd, &cmd);
+				break;
+			default:
+				fprintf(stderr, "%s(): invalid message "
+				    "received.\n", __func__);
+				error = EINVAL;
+				goto out;
+		}
+	}
+
+out:
+	errno = error;
+	return error ? -1 : 0;
+}
--- a/zfs/lib/libdmu-ctl/dctl_common.c
+++ b/zfs/lib/libdmu-ctl/dctl_common.c
@ -0,0 +1,109 @@
+/*
+ * CDDL HEADER START
+ *
+ * The contents of this file are subject to the terms of the
+ * Common Development and Distribution License, Version 1.0 only
+ * (the "License").  You may not use this file except in compliance
+ * with the License.
+ *
+ * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
+ * or http://www.opensolaris.org/os/licensing.
+ * See the License for the specific language governing permissions
+ * and limitations under the License.
+ *
+ * When distributing Covered Code, include this CDDL HEADER in each
+ * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
+ * If applicable, add the following below this CDDL HEADER, with the
+ * fields enclosed by brackets "[]" replaced with your own identifying
+ * information: Portions Copyright [yyyy] [name of copyright owner]
+ *
+ * CDDL HEADER END
+ */
+/*
+ * Copyright 2008 Sun Microsystems, Inc.  All rights reserved.
+ * Use is subject to license terms.
+ */
+
+#include <stdio.h>
+#include <errno.h>
+#include <sys/types.h>
+#include <sys/socket.h>
+
+#include <sys/dmu_ctl.h>
+#include <sys/dmu_ctl_impl.h>
+
+int dctl_read_msg(int fd, dctl_cmd_t *cmd)
+{
+	int error;
+
+	/*
+	 * First, read only the magic number and the protocol version.
+	 *
+	 * This prevents blocking forever in case the size of dctl_cmd_t
+	 * shrinks in future protocol versions.
+	 */
+	error = dctl_read_data(fd, cmd, DCTL_CMD_HEADER_SIZE);
+
+	if (!error &&cmd->dcmd_magic != DCTL_MAGIC) {
+		fprintf(stderr, "%s(): invalid magic number\n", __func__);
+		error = EIO;
+	}
+
+	if (!error && cmd->dcmd_version != DCTL_PROTOCOL_VER) {
+		fprintf(stderr, "%s(): invalid protocol version\n", __func__);
+		error = ENOTSUP;
+	}
+
+	if (error)
+		return error;
+
+	/* Get the rest of the command */
+	return dctl_read_data(fd, (caddr_t) cmd + DCTL_CMD_HEADER_SIZE,
+	    sizeof(dctl_cmd_t) - DCTL_CMD_HEADER_SIZE);
+}
+
+int dctl_send_msg(int fd, dctl_cmd_t *cmd)
+{
+	cmd->dcmd_magic = DCTL_MAGIC;
+	cmd->dcmd_version = DCTL_PROTOCOL_VER;
+
+	return dctl_send_data(fd, cmd, sizeof(dctl_cmd_t));
+}
+
+int dctl_read_data(int fd, void *ptr, size_t size)
+{
+	size_t read = 0;
+	size_t left = size;
+	ssize_t rc;
+
+	while (left > 0) {
+		rc = recv(fd, (caddr_t) ptr + read, left, 0);
+
+		/* File descriptor closed */
+		if (rc == 0)
+			return ECONNRESET;
+
+		if (rc == -1) {
+			if (errno == EINTR)
+				continue;
+			return errno;
+		}
+
+		read += rc;
+		left -= rc;
+	}
+
+	return 0;
+}
+
+int dctl_send_data(int fd, const void *ptr, size_t size)
+{
+	ssize_t rc;
+
+	do {
+		rc = send(fd, ptr, size, MSG_NOSIGNAL);
+	} while(rc == -1 && errno == EINTR);
+
+	return rc == size ? 0 : EIO;
+}
+
--- a/zfs/lib/libdmu-ctl/dctl_server.c
+++ b/zfs/lib/libdmu-ctl/dctl_server.c
@ -0,0 +1,476 @@
+/*
+ * CDDL HEADER START
+ *
+ * The contents of this file are subject to the terms of the
+ * Common Development and Distribution License, Version 1.0 only
+ * (the "License").  You may not use this file except in compliance
+ * with the License.
+ *
+ * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
+ * or http://www.opensolaris.org/os/licensing.
+ * See the License for the specific language governing permissions
+ * and limitations under the License.
+ *
+ * When distributing Covered Code, include this CDDL HEADER in each
+ * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
+ * If applicable, add the following below this CDDL HEADER, with the
+ * fields enclosed by brackets "[]" replaced with your own identifying
+ * information: Portions Copyright [yyyy] [name of copyright owner]
+ *
+ * CDDL HEADER END
+ */
+/*
+ * Copyright 2008 Sun Microsystems, Inc.  All rights reserved.
+ * Use is subject to license terms.
+ */
+
+#include <stdio.h>
+#include <stddef.h>
+#include <stdlib.h>
+#include <string.h>
+#include <signal.h>
+#include <limits.h>
+#include <errno.h>
+#include <poll.h>
+#include <pthread.h>
+#include <unistd.h>
+#include <sys/debug.h>
+#include <sys/socket.h>
+#include <sys/stat.h>
+#include <sys/types.h>
+#include <sys/un.h>
+#include <sys/list.h>
+#include <sys/cred.h>
+
+#include <sys/dmu_ctl.h>
+#include <sys/dmu_ctl_impl.h>
+
+static dctl_sock_info_t ctl_sock = {
+	.dsi_mtx = PTHREAD_MUTEX_INITIALIZER,
+	.dsi_fd = -1
+};
+
+static int dctl_create_socket_common();
+
+/*
+ * Routines from zfs_ioctl.c
+ */
+extern int zfs_ioctl_init();
+extern int zfs_ioctl_fini();
+extern int zfsdev_ioctl(dev_t dev, int cmd, intptr_t arg, int flag, cred_t *cr,
+    int *rvalp);
+
+/*
+ * We can't simply put the client file descriptor in wthr_info_t because we
+ * have no way of accessing it from the DMU code without extensive
+ * modifications.
+ *
+ * Therefore each worker thread will have it's own global thread-specific
+ * client_fd variable.
+ */
+static __thread int client_fd = -1;
+
+int dctls_copyin(const void *src, void *dest, size_t size)
+{
+	dctl_cmd_t cmd;
+
+	VERIFY(client_fd >= 0);
+
+	cmd.dcmd_msg = DCTL_COPYIN;
+	cmd.u.dcmd_copy.ptr = (uintptr_t) src;
+	cmd.u.dcmd_copy.size = size;
+
+	if (dctl_send_msg(client_fd, &cmd) != 0)
+		return EFAULT;
+
+	if (dctl_read_data(client_fd, dest, size) != 0)
+		return EFAULT;
+
+	return 0;
+}
+
+int dctls_copyinstr(const char *from, char *to, size_t max, size_t *len)
+{
+	dctl_cmd_t msg;
+	size_t copied;
+
+	VERIFY(client_fd >= 0);
+
+	if (max == 0)
+		return ENAMETOOLONG;
+	if (max < 0)
+		return EFAULT;
+
+	msg.dcmd_msg = DCTL_COPYINSTR;
+	msg.u.dcmd_copy.ptr = (uintptr_t) from;
+	msg.u.dcmd_copy.size = max;
+
+	if (dctl_send_msg(client_fd, &msg) != 0)
+		return EFAULT;
+
+	if (dctl_read_msg(client_fd, &msg) != 0)
+		return EFAULT;
+
+	if (msg.dcmd_msg != DCTL_GEN_REPLY)
+		return EFAULT;
+
+	copied = msg.u.dcmd_reply.size;
+
+	if (copied >= max)
+		return EFAULT;
+
+	if (copied > 0)
+		if (dctl_read_data(client_fd, to, copied) != 0)
+			return EFAULT;
+
+	to[copied] = '\0';
+
+	if (len != NULL)
+		*len = copied + 1;
+
+	return msg.u.dcmd_reply.rc;
+}
+
+int dctls_copyout(const void *src, void *dest, size_t size)
+{
+	dctl_cmd_t cmd;
+
+	VERIFY(client_fd >= 0);
+
+	cmd.dcmd_msg = DCTL_COPYOUT;
+	cmd.u.dcmd_copy.ptr = (uintptr_t) dest;
+	cmd.u.dcmd_copy.size = size;
+
+	if (dctl_send_msg(client_fd, &cmd) != 0)
+		return EFAULT;
+
+	if (dctl_send_data(client_fd, src, size) != 0)
+		return EFAULT;
+
+	return 0;
+}
+
+int dctls_fd_read(int fd, void *buf, ssize_t len, ssize_t *residp)
+{
+	dctl_cmd_t msg;
+	uint64_t dsize;
+	int error;
+
+	VERIFY(client_fd >= 0);
+
+	msg.dcmd_msg = DCTL_FD_READ;
+	msg.u.dcmd_fd_io.fd = fd;
+	msg.u.dcmd_fd_io.size = len;
+
+	if ((error = dctl_send_msg(client_fd, &msg)) != 0)
+		return error;
+
+	if ((error = dctl_read_msg(client_fd, &msg)) != 0)
+		return error;
+
+	if (msg.dcmd_msg != DCTL_GEN_REPLY)
+		return EIO;
+
+	if (msg.u.dcmd_reply.rc != 0)
+		return msg.u.dcmd_reply.rc;
+
+	dsize = msg.u.dcmd_reply.size;
+
+	if (dsize > 0)
+		error = dctl_read_data(client_fd, buf, dsize);
+
+	*residp = len - dsize;
+
+	return error;
+}
+
+int dctls_fd_write(int fd, const void *src, ssize_t len)
+{
+	dctl_cmd_t msg;
+	int error;
+
+	VERIFY(client_fd >= 0);
+
+	msg.dcmd_msg = DCTL_FD_WRITE;
+	msg.u.dcmd_fd_io.fd = fd;
+	msg.u.dcmd_fd_io.size = len;
+
+	error = dctl_send_msg(client_fd, &msg);
+
+	if (!error)
+		error = dctl_send_data(client_fd, src, len);
+
+	if (!error)
+		error = dctl_read_msg(client_fd, &msg);
+
+	if (error)
+		return error;
+
+	if (msg.dcmd_msg != DCTL_GEN_REPLY)
+		return EIO;
+
+	if (msg.u.dcmd_reply.rc != 0)
+		return msg.u.dcmd_reply.rc;
+
+	/*
+	 * We have to do this because the original upstream code
+	 * does not check if residp == len.
+	 */
+	if (msg.u.dcmd_reply.size != len)
+		return EIO;
+
+	return 0;
+}
+
+/* Handle a new connection */
+static void dctl_handle_conn(int sock_fd)
+{
+	dctl_cmd_t cmd;
+	dev_t dev = { 0 };
+	int rc;
+
+	client_fd = sock_fd;
+
+	while (dctl_read_msg(sock_fd, &cmd) == 0) {
+		if (cmd.dcmd_msg != DCTL_IOCTL) {
+			fprintf(stderr, "%s(): unexpected message type.\n",
+			    __func__);
+			break;
+		}
+
+		rc = zfsdev_ioctl(dev, cmd.u.dcmd_ioctl.cmd,
+		    (intptr_t) cmd.u.dcmd_ioctl.arg, 0, NULL, NULL);
+
+		cmd.dcmd_msg = DCTL_IOCTL_REPLY;
+		cmd.u.dcmd_reply.rc = rc;
+
+		if (dctl_send_msg(sock_fd, &cmd) != 0)
+			break;
+	}
+	close(sock_fd);
+
+	client_fd = -1;
+}
+
+/* Main worker thread loop */
+static void *dctl_thread(void *arg)
+{
+	wthr_info_t *thr = arg;
+	struct pollfd fds[1];
+
+	fds[0].events = POLLIN;
+
+	pthread_mutex_lock(&ctl_sock.dsi_mtx);
+
+	while (!thr->wthr_exit) {
+		/* Clean-up dead threads */
+		dctl_thr_join();
+
+		/* The file descriptor might change in the thread lifetime */
+		fds[0].fd = ctl_sock.dsi_fd;
+
+		/* Poll socket with 1-second timeout */
+		int rc = poll(fds, 1, 1000);
+		if (rc == 0 || (rc == -1 && errno == EINTR))
+			continue;
+
+		/* Recheck the exit flag */
+		if (thr->wthr_exit)
+			break;
+
+		if (rc == -1) {
+			/* Unknown error, let's try to recreate the socket */
+			close(ctl_sock.dsi_fd);
+			ctl_sock.dsi_fd = -1;
+
+			if (dctl_create_socket_common() != 0)
+				break;
+
+			continue;
+		}
+		ASSERT(rc == 1);
+
+		short rev = fds[0].revents;
+		if (rev == 0)
+			continue;
+		ASSERT(rev == POLLIN);
+
+		/*
+		 * At this point there should be a connection ready to be
+		 * accepted.
+		 */
+		int client_fd = accept(ctl_sock.dsi_fd, NULL, NULL);
+		/* Many possible errors here, we'll just retry */
+		if (client_fd == -1)
+			continue;
+
+		/*
+		 * Now lets handle the request. This can take a very
+		 * long time (hours even), so we'll let other threads
+		 * handle new connections.
+		 */
+		pthread_mutex_unlock(&ctl_sock.dsi_mtx);
+
+		dctl_thr_rebalance(thr, B_FALSE);
+		dctl_handle_conn(client_fd);
+		dctl_thr_rebalance(thr, B_TRUE);
+
+		pthread_mutex_lock(&ctl_sock.dsi_mtx);
+	}
+	pthread_mutex_unlock(&ctl_sock.dsi_mtx);
+
+	dctl_thr_die(thr);
+
+	return NULL;
+}
+
+static int dctl_create_socket_common()
+{
+	dctl_sock_info_t *s = &ctl_sock;
+	size_t size;
+	int error;
+
+	ASSERT(s->dsi_fd == -1);
+
+	/*
+	 * Unlink old socket, in case it exists.
+	 * We don't care about errors here.
+	 */
+	unlink(s->dsi_path);
+
+	/* Create the socket */
+	s->dsi_fd = socket(PF_UNIX, SOCK_STREAM, 0);
+	if (s->dsi_fd == -1) {
+		error = errno;
+		perror("socket");
+		return error;
+	}
+
+	s->dsi_addr.sun_family = AF_UNIX;
+
+	size = sizeof(s->dsi_addr.sun_path) - 1;
+	strncpy(s->dsi_addr.sun_path, s->dsi_path, size);
+
+	s->dsi_addr.sun_path[size] = '\0';
+
+	if (bind(s->dsi_fd, (struct sockaddr *) &s->dsi_addr,
+	    sizeof(s->dsi_addr)) != 0) {
+		error = errno;
+		perror("bind");
+		return error;
+	}
+
+	if (listen(s->dsi_fd, LISTEN_BACKLOG) != 0) {
+		error = errno;
+		perror("listen");
+		unlink(s->dsi_path);
+		return error;
+	}
+
+	return 0;
+}
+
+static int dctl_create_socket(const char *cfg_dir)
+{
+	int error;
+	dctl_sock_info_t *s = &ctl_sock;
+
+	ASSERT(s->dsi_path == NULL);
+	ASSERT(s->dsi_fd == -1);
+
+	int pathsize = strlen(cfg_dir) + strlen(SOCKNAME) + 2;
+	if (pathsize > sizeof(s->dsi_addr.sun_path))
+		return ENAMETOOLONG;
+
+	s->dsi_path = malloc(pathsize);
+	if (s->dsi_path == NULL)
+		return ENOMEM;
+
+	strcpy(s->dsi_path, cfg_dir);
+	strcat(s->dsi_path, "/" SOCKNAME);
+
+	/*
+	 * For convenience, create the directory in case it doesn't exist.
+	 * We don't care about errors here.
+	 */
+	mkdir(cfg_dir, 0770);
+
+	error = dctl_create_socket_common();
+
+	if (error) {
+		free(s->dsi_path);
+
+		if (s->dsi_fd != -1) {
+			close(s->dsi_fd);
+			s->dsi_fd = -1;
+		}
+	}
+
+	return error;
+}
+
+static void dctl_destroy_socket()
+{
+	dctl_sock_info_t *s = &ctl_sock;
+
+	ASSERT(s->dsi_path != NULL);
+	ASSERT(s->dsi_fd != -1);
+
+	close(s->dsi_fd);
+	s->dsi_fd = -1;
+
+	unlink(s->dsi_path);
+	free(s->dsi_path);
+}
+
+/*
+ * Initialize the DMU userspace control interface.
+ * This should be called after kernel_init().
+ *
+ * Note that only very rarely we have more than a couple of simultaneous
+ * lzfs/lzpool connections. Since the thread pool grows automatically when all
+ * threads are busy, a good value for min_thr and max_free_thr is 2.
+ */
+int dctl_server_init(const char *cfg_dir, int min_thr, int max_free_thr)
+{
+	int error;
+
+	ASSERT(min_thr > 0);
+	ASSERT(max_free_thr >= min_thr);
+
+	error = zfs_ioctl_init();
+	if (error)
+		return error;
+
+	error = dctl_create_socket(cfg_dir);
+	if (error) {
+		(void) zfs_ioctl_fini();
+		return error;
+	}
+
+	error = dctl_thr_pool_create(min_thr, max_free_thr, dctl_thread);
+	if (error) {
+		(void) zfs_ioctl_fini();
+		dctl_destroy_socket();
+		return error;
+	}
+
+	return 0;
+}
+
+/*
+ * Terminate control interface.
+ * This should be called after closing all objsets, but before calling
+ * kernel_fini().
+ * May return EBUSY if the SPA is busy.
+ *
+ * Thread pool destruction can take a while due to poll()
+ * timeout or due to a thread being busy (e.g. a backup is being taken).
+ */
+int dctl_server_fini()
+{
+	dctl_thr_pool_stop();
+	dctl_destroy_socket();
+
+	return zfs_ioctl_fini();
+}
--- a/zfs/lib/libdmu-ctl/dctl_thrpool.c
+++ b/zfs/lib/libdmu-ctl/dctl_thrpool.c
@ -0,0 +1,253 @@
+/*
+ * CDDL HEADER START
+ *
+ * The contents of this file are subject to the terms of the
+ * Common Development and Distribution License, Version 1.0 only
+ * (the "License").  You may not use this file except in compliance
+ * with the License.
+ *
+ * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
+ * or http://www.opensolaris.org/os/licensing.
+ * See the License for the specific language governing permissions
+ * and limitations under the License.
+ *
+ * When distributing Covered Code, include this CDDL HEADER in each
+ * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
+ * If applicable, add the following below this CDDL HEADER, with the
+ * fields enclosed by brackets "[]" replaced with your own identifying
+ * information: Portions Copyright [yyyy] [name of copyright owner]
+ *
+ * CDDL HEADER END
+ */
+/*
+ * Copyright 2008 Sun Microsystems, Inc.  All rights reserved.
+ * Use is subject to license terms.
+ */
+
+#include <stdlib.h>
+#include <stddef.h>
+#include <time.h>
+#include <pthread.h>
+#include <errno.h>
+#include <sys/list.h>
+#include <sys/debug.h>
+
+#include <sys/dmu_ctl.h>
+#include <sys/dmu_ctl_impl.h>
+
+static dctl_thr_info_t thr_pool = {
+	.dti_mtx = PTHREAD_MUTEX_INITIALIZER
+};
+
+/*
+ * Create n threads.
+ * Callers must acquire thr_pool.dti_mtx first.
+ */
+static int dctl_thr_create(int n)
+{
+	dctl_thr_info_t *p = &thr_pool;
+	int error;
+
+	for (int i = 0; i < n; i++) {
+		wthr_info_t *thr = malloc(sizeof(wthr_info_t));
+		if (thr == NULL)
+			return ENOMEM;
+
+		thr->wthr_exit = B_FALSE;
+		thr->wthr_free = B_TRUE;
+
+		error = pthread_create(&thr->wthr_id, NULL, p->dti_thr_func,
+		    thr);
+		if (error) {
+			free(thr);
+			return error;
+		}
+
+		p->dti_free++;
+
+		list_insert_tail(&p->dti_list, thr);
+	}
+	return 0;
+}
+
+/*
+ * Mark the thread as dead.
+ * Must be called right before exiting the main thread function.
+ */
+void dctl_thr_die(wthr_info_t *thr)
+{
+	dctl_thr_info_t *p = &thr_pool;
+
+	thr->wthr_exit = B_TRUE;
+	dctl_thr_rebalance(thr, B_FALSE);
+
+	pthread_mutex_lock(&p->dti_mtx);
+
+	list_remove(&p->dti_list, thr);
+	list_insert_tail(&p->dti_join_list, thr);
+
+	pthread_mutex_unlock(&p->dti_mtx);
+}
+
+/*
+ * Clean-up dead threads.
+ */
+void dctl_thr_join()
+{
+	dctl_thr_info_t *p = &thr_pool;
+	wthr_info_t *thr;
+
+	pthread_mutex_lock(&p->dti_mtx);
+
+	while ((thr = list_head(&p->dti_join_list))) {
+		list_remove(&p->dti_join_list, thr);
+
+		ASSERT(!pthread_equal(thr->wthr_id, pthread_self()));
+
+		/*
+		 * This should not block because all the threads
+		 * on this list should have died already.
+		 *
+		 * pthread_join() can only return an error if
+		 * we made a programming mistake.
+		 */
+		VERIFY(pthread_join(thr->wthr_id, NULL) == 0);
+
+		ASSERT(thr->wthr_exit);
+		ASSERT(!thr->wthr_free);
+
+		free(thr);
+	}
+
+	pthread_mutex_unlock(&p->dti_mtx);
+}
+
+/*
+ * Adjust the number of free threads in the pool and the thread status.
+ *
+ * Callers must acquire thr_pool.dti_mtx first.
+ */
+static void dctl_thr_adjust_free(wthr_info_t *thr, boolean_t set_free)
+{
+	dctl_thr_info_t *p = &thr_pool;
+
+	ASSERT(p->dti_free >= 0);
+
+	if (!thr->wthr_free && set_free)
+		p->dti_free++;
+	else if (thr->wthr_free && !set_free)
+		p->dti_free--;
+
+	ASSERT(p->dti_free >= 0);
+
+	thr->wthr_free = set_free;
+}
+
+/*
+ * Rebalance threads. Also adjusts the free status of the thread.
+ * Will set the thread exit flag if the number of free threads is above
+ * the limit.
+ */
+void dctl_thr_rebalance(wthr_info_t *thr, boolean_t set_free)
+{
+	dctl_thr_info_t *p = &thr_pool;
+
+	pthread_mutex_lock(&p->dti_mtx);
+
+	if (p->dti_exit || p->dti_free > p->dti_max_free)
+		thr->wthr_exit = B_TRUE;
+
+	if (thr->wthr_exit)
+		set_free = B_FALSE;
+
+	dctl_thr_adjust_free(thr, set_free);
+
+	if (!p->dti_exit && p->dti_free == 0)
+		dctl_thr_create(1);
+
+	pthread_mutex_unlock(&p->dti_mtx);
+}
+
+/*
+ * Stop the thread pool.
+ *
+ * This can take a while since it actually waits for all threads to exit.
+ */
+void dctl_thr_pool_stop()
+{
+	dctl_thr_info_t *p = &thr_pool;
+	wthr_info_t *thr;
+	struct timespec ts;
+
+	pthread_mutex_lock(&p->dti_mtx);
+
+	ASSERT(!p->dti_exit);
+	p->dti_exit = B_TRUE;
+
+	/* Let's flag the threads first */
+	thr = list_head(&p->dti_list);
+	while (thr != NULL) {
+		thr->wthr_exit = B_TRUE;
+		dctl_thr_adjust_free(thr, B_FALSE);
+
+		thr = list_next(&p->dti_list, thr);
+	}
+
+	pthread_mutex_unlock(&p->dti_mtx);
+
+	/* Now let's wait for them to exit */
+	ts.tv_sec = 0;
+	ts.tv_nsec = 50000000; /* 50ms */
+	do {
+		nanosleep(&ts, NULL);
+
+		pthread_mutex_lock(&p->dti_mtx);
+		thr = list_head(&p->dti_list);
+		pthread_mutex_unlock(&p->dti_mtx);
+
+		dctl_thr_join();
+	} while(thr != NULL);
+
+	ASSERT(p->dti_free == 0);
+
+	ASSERT(list_is_empty(&p->dti_list));
+	ASSERT(list_is_empty(&p->dti_join_list));
+
+	list_destroy(&p->dti_list);
+	list_destroy(&p->dti_join_list);
+}
+
+/*
+ * Create thread pool.
+ *
+ * If at least one thread creation fails, it will stop all previous
+ * threads and return a non-zero value.
+ */
+int dctl_thr_pool_create(int min_thr, int max_free_thr,
+    thr_func_t *thr_func)
+{
+	int error;
+	dctl_thr_info_t *p = &thr_pool;
+
+	ASSERT(p->dti_free == 0);
+
+	/* Initialize global variables */
+	p->dti_min = min_thr;
+	p->dti_max_free = max_free_thr;
+	p->dti_exit = B_FALSE;
+	p->dti_thr_func = thr_func;
+
+	list_create(&p->dti_list, sizeof(wthr_info_t), offsetof(wthr_info_t,
+	    wthr_node));
+	list_create(&p->dti_join_list, sizeof(wthr_info_t),
+	    offsetof(wthr_info_t, wthr_node));
+
+	pthread_mutex_lock(&p->dti_mtx);
+	error = dctl_thr_create(min_thr);
+	pthread_mutex_unlock(&p->dti_mtx);
+
+	if (error)
+		dctl_thr_pool_stop();
+
+	return error;
+}
--- a/zfs/lib/libdmu-ctl/dmu_send.c
+++ b/zfs/lib/libdmu-ctl/dmu_send.c
--- a/zfs/lib/libdmu-ctl/include/Makefile.in
+++ b/zfs/lib/libdmu-ctl/include/Makefile.in
@ -0,0 +1 @@
+subdir-m += sys
--- a/zfs/lib/libdmu-ctl/include/sys/Makefile.in
+++ b/zfs/lib/libdmu-ctl/include/sys/Makefile.in
@ -0,0 +1 @@
+DISTFILES = dmu_ctl.h dmu_ctl_impl.h
--- a/zfs/lib/libdmu-ctl/include/sys/dmu_ctl.h
+++ b/zfs/lib/libdmu-ctl/include/sys/dmu_ctl.h
@ -0,0 +1,71 @@
+/*
+ * CDDL HEADER START
+ *
+ * The contents of this file are subject to the terms of the
+ * Common Development and Distribution License, Version 1.0 only
+ * (the "License").  You may not use this file except in compliance
+ * with the License.
+ *
+ * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
+ * or http://www.opensolaris.org/os/licensing.
+ * See the License for the specific language governing permissions
+ * and limitations under the License.
+ *
+ * When distributing Covered Code, include this CDDL HEADER in each
+ * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
+ * If applicable, add the following below this CDDL HEADER, with the
+ * fields enclosed by brackets "[]" replaced with your own identifying
+ * information: Portions Copyright [yyyy] [name of copyright owner]
+ *
+ * CDDL HEADER END
+ */
+/*
+ * Copyright 2008 Sun Microsystems, Inc.  All rights reserved.
+ * Use is subject to license terms.
+ */
+
+#ifndef _SYS_DMU_CTL_H
+#define _SYS_DMU_CTL_H
+
+#include <sys/types.h>
+
+/* Default directory where the clients search for sockets to connect */
+#define DMU_CTL_DEFAULT_DIR "/var/run/zfs/udmu"
+
+/*
+ * These functions are called by the server process.
+ *
+ * kernel_init() must be called before dctl_server_init().
+ * kernel_fini() must not be called before dctl_server_fini().
+ *
+ * All objsets must be closed and object references be released before calling
+ * dctl_server_fini(), otherwise it will return EBUSY.
+ *
+ * Note: On Solaris, it is highly recommended to either catch or ignore the
+ * SIGPIPE signal, otherwise the server process will die if the client is
+ * killed.
+ */
+int dctl_server_init(const char *cfg_dir, int min_threads,
+    int max_free_threads);
+int dctl_server_fini();
+
+/*
+ * The following functions are called by the DMU from the server process context
+ * (in the worker threads).
+ */
+int dctls_copyin(const void *src, void *dest, size_t size);
+int dctls_copyinstr(const char *from, char *to, size_t max,
+    size_t *len);
+int dctls_copyout(const void *src, void *dest, size_t size);
+int dctls_fd_read(int fd, void *buf, ssize_t len, ssize_t *residp);
+int dctls_fd_write(int fd, const void *src, ssize_t len);
+
+/*
+ * These functions are called by the client process (libzfs).
+ */
+int dctlc_connect(const char *dir, boolean_t check_subdirs);
+void dctlc_disconnect(int fd);
+
+int dctlc_ioctl(int fd, int32_t request, void *arg);
+
+#endif
--- a/zfs/lib/libdmu-ctl/include/sys/dmu_ctl_impl.h
+++ b/zfs/lib/libdmu-ctl/include/sys/dmu_ctl_impl.h
@ -0,0 +1,144 @@
+/*
+ * CDDL HEADER START
+ *
+ * The contents of this file are subject to the terms of the
+ * Common Development and Distribution License, Version 1.0 only
+ * (the "License").  You may not use this file except in compliance
+ * with the License.
+ *
+ * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
+ * or http://www.opensolaris.org/os/licensing.
+ * See the License for the specific language governing permissions
+ * and limitations under the License.
+ *
+ * When distributing Covered Code, include this CDDL HEADER in each
+ * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
+ * If applicable, add the following below this CDDL HEADER, with the
+ * fields enclosed by brackets "[]" replaced with your own identifying
+ * information: Portions Copyright [yyyy] [name of copyright owner]
+ *
+ * CDDL HEADER END
+ */
+/*
+ * Copyright 2008 Sun Microsystems, Inc.  All rights reserved.
+ * Use is subject to license terms.
+ */
+
+#ifndef _SYS_DMU_CTL_IMPL_H
+#define _SYS_DMU_CTL_IMPL_H
+
+#include <sys/list.h>
+#include <sys/types.h>
+#include <sys/socket.h>
+#include <sys/un.h>
+#include <pthread.h>
+
+#define SOCKNAME "dmu_socket"
+
+#define DCTL_PROTOCOL_VER 1
+#define DCTL_MAGIC 0xdc71b1070c01dc71ll
+
+/* Message types */
+enum {
+	DCTL_IOCTL,
+	DCTL_IOCTL_REPLY,
+	DCTL_COPYIN,
+	DCTL_COPYINSTR,
+	DCTL_COPYOUT,
+	DCTL_FD_READ,
+	DCTL_FD_WRITE,
+	DCTL_GEN_REPLY /* generic reply */
+};
+
+/* On-the-wire message */
+typedef struct dctl_cmd {
+	uint64_t dcmd_magic;
+	int8_t   dcmd_version;
+	int8_t   dcmd_msg;
+	uint8_t  dcmd_pad[6];
+	union {
+		struct dcmd_ioctl {
+			uint64_t arg;
+			int32_t cmd;
+			uint8_t pad[4];
+		} dcmd_ioctl;
+
+		struct dcmd_copy_req {
+			uint64_t ptr;
+			uint64_t size;
+		} dcmd_copy;
+
+		struct dcmd_fd_req {
+			int64_t size;
+			int32_t fd;
+			uint8_t pad[4];
+		} dcmd_fd_io;
+
+		struct dcmd_reply {
+			uint64_t size;  /* used by reply to DCTL_COPYINSTR,
+			                   DCTL_FD_READ and DCTL_FD_WRITE */
+			int32_t rc;     /* return code */
+			uint8_t pad[4];
+		} dcmd_reply;
+	} u;
+} dctl_cmd_t;
+
+#define DCTL_CMD_HEADER_SIZE (sizeof(uint64_t) + sizeof(uint8_t))
+
+/*
+ * The following definitions are only used by the server code.
+ */
+
+#define LISTEN_BACKLOG 5
+
+/* Worker thread data */
+typedef struct wthr_info {
+	list_node_t wthr_node;
+	pthread_t   wthr_id;
+	boolean_t   wthr_exit; /* termination flag */
+	boolean_t   wthr_free;
+} wthr_info_t;
+
+/* Control socket data */
+typedef struct dctl_sock_info {
+	pthread_mutex_t    dsi_mtx;
+	char               *dsi_path;
+	struct sockaddr_un dsi_addr;
+	int                dsi_fd;
+} dctl_sock_info_t;
+
+typedef void *thr_func_t(void *);
+
+/* Thread pool data */
+typedef struct dctl_thr_info {
+	thr_func_t *dti_thr_func;
+
+	pthread_mutex_t dti_mtx; /* protects the thread lists and dti_free */
+	list_t dti_list;         /* list of threads in the thread pool */
+	list_t dti_join_list;    /* list of threads that are waiting to be
+	                            joined */
+	int    dti_free;         /* number of free worker threads */
+
+	int dti_min;
+	int dti_max_free;
+
+	boolean_t dti_exit; /* global termination flag */
+} dctl_thr_info_t;
+
+/* Messaging functions functions */
+int dctl_read_msg(int fd, dctl_cmd_t *cmd);
+int dctl_send_msg(int fd, dctl_cmd_t *cmd);
+
+int dctl_read_data(int fd, void *ptr, size_t size);
+int dctl_send_data(int fd, const void *ptr, size_t size);
+
+/* Thread pool functions */
+int dctl_thr_pool_create(int min_thr, int max_free_thr,
+    thr_func_t *thr_func);
+void dctl_thr_pool_stop();
+
+void dctl_thr_join();
+void dctl_thr_die(wthr_info_t *thr);
+void dctl_thr_rebalance(wthr_info_t *thr, boolean_t set_free);
+
+#endif
--- a/zfs/lib/libdmu-ctl/rrwlock.c
+++ b/zfs/lib/libdmu-ctl/rrwlock.c
@ -0,0 +1,249 @@
+/*
+ * CDDL HEADER START
+ *
+ * The contents of this file are subject to the terms of the
+ * Common Development and Distribution License (the "License").
+ * You may not use this file except in compliance with the License.
+ *
+ * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
+ * or http://www.opensolaris.org/os/licensing.
+ * See the License for the specific language governing permissions
+ * and limitations under the License.
+ *
+ * When distributing Covered Code, include this CDDL HEADER in each
+ * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
+ * If applicable, add the following below this CDDL HEADER, with the
+ * fields enclosed by brackets "[]" replaced with your own identifying
+ * information: Portions Copyright [yyyy] [name of copyright owner]
+ *
+ * CDDL HEADER END
+ */
+/*
+ * Copyright 2007 Sun Microsystems, Inc.  All rights reserved.
+ * Use is subject to license terms.
+ */
+
+#pragma ident	"@(#)rrwlock.c	1.1	07/10/24 SMI"
+
+#include <sys/refcount.h>
+#include <sys/rrwlock.h>
+
+/*
+ * This file contains the implementation of a re-entrant read
+ * reader/writer lock (aka "rrwlock").
+ *
+ * This is a normal reader/writer lock with the additional feature
+ * of allowing threads who have already obtained a read lock to
+ * re-enter another read lock (re-entrant read) - even if there are
+ * waiting writers.
+ *
+ * Callers who have not obtained a read lock give waiting writers priority.
+ *
+ * The rrwlock_t lock does not allow re-entrant writers, nor does it
+ * allow a re-entrant mix of reads and writes (that is, it does not
+ * allow a caller who has already obtained a read lock to be able to
+ * then grab a write lock without first dropping all read locks, and
+ * vice versa).
+ *
+ * The rrwlock_t uses tsd (thread specific data) to keep a list of
+ * nodes (rrw_node_t), where each node keeps track of which specific
+ * lock (rrw_node_t::rn_rrl) the thread has grabbed.  Since re-entering
+ * should be rare, a thread that grabs multiple reads on the same rrwlock_t
+ * will store multiple rrw_node_ts of the same 'rrn_rrl'. Nodes on the
+ * tsd list can represent a different rrwlock_t.  This allows a thread
+ * to enter multiple and unique rrwlock_ts for read locks at the same time.
+ *
+ * Since using tsd exposes some overhead, the rrwlock_t only needs to
+ * keep tsd data when writers are waiting.  If no writers are waiting, then
+ * a reader just bumps the anonymous read count (rr_anon_rcount) - no tsd
+ * is needed.  Once a writer attempts to grab the lock, readers then
+ * keep tsd data and bump the linked readers count (rr_linked_rcount).
+ *
+ * If there are waiting writers and there are anonymous readers, then a
+ * reader doesn't know if it is a re-entrant lock. But since it may be one,
+ * we allow the read to proceed (otherwise it could deadlock).  Since once
+ * waiting writers are active, readers no longer bump the anonymous count,
+ * the anonymous readers will eventually flush themselves out.  At this point,
+ * readers will be able to tell if they are a re-entrant lock (have a
+ * rrw_node_t entry for the lock) or not. If they are a re-entrant lock, then
+ * we must let the proceed.  If they are not, then the reader blocks for the
+ * waiting writers.  Hence, we do not starve writers.
+ */
+
+/* global key for TSD */
+uint_t rrw_tsd_key;
+
+typedef struct rrw_node {
+	struct rrw_node	*rn_next;
+	rrwlock_t	*rn_rrl;
+} rrw_node_t;
+
+static rrw_node_t *
+rrn_find(rrwlock_t *rrl)
+{
+	rrw_node_t *rn;
+
+	if (refcount_count(&rrl->rr_linked_rcount) == 0)
+		return (NULL);
+
+	for (rn = tsd_get(rrw_tsd_key); rn != NULL; rn = rn->rn_next) {
+		if (rn->rn_rrl == rrl)
+			return (rn);
+	}
+	return (NULL);
+}
+
+/*
+ * Add a node to the head of the singly linked list.
+ */
+static void
+rrn_add(rrwlock_t *rrl)
+{
+	rrw_node_t *rn;
+
+	rn = kmem_alloc(sizeof (*rn), KM_SLEEP);
+	rn->rn_rrl = rrl;
+	rn->rn_next = tsd_get(rrw_tsd_key);
+	VERIFY(tsd_set(rrw_tsd_key, rn) == 0);
+}
+
+/*
+ * If a node is found for 'rrl', then remove the node from this
+ * thread's list and return TRUE; otherwise return FALSE.
+ */
+static boolean_t
+rrn_find_and_remove(rrwlock_t *rrl)
+{
+	rrw_node_t *rn;
+	rrw_node_t *prev = NULL;
+
+	if (refcount_count(&rrl->rr_linked_rcount) == 0)
+		return (NULL);
+
+	for (rn = tsd_get(rrw_tsd_key); rn != NULL; rn = rn->rn_next) {
+		if (rn->rn_rrl == rrl) {
+			if (prev)
+				prev->rn_next = rn->rn_next;
+			else
+				VERIFY(tsd_set(rrw_tsd_key, rn->rn_next) == 0);
+			kmem_free(rn, sizeof (*rn));
+			return (B_TRUE);
+		}
+		prev = rn;
+	}
+	return (B_FALSE);
+}
+
+void
+rrw_init(rrwlock_t *rrl)
+{
+	mutex_init(&rrl->rr_lock, NULL, MUTEX_DEFAULT, NULL);
+	cv_init(&rrl->rr_cv, NULL, CV_DEFAULT, NULL);
+	rrl->rr_writer = NULL;
+	refcount_create(&rrl->rr_anon_rcount);
+	refcount_create(&rrl->rr_linked_rcount);
+	rrl->rr_writer_wanted = B_FALSE;
+}
+
+void
+rrw_destroy(rrwlock_t *rrl)
+{
+	mutex_destroy(&rrl->rr_lock);
+	cv_destroy(&rrl->rr_cv);
+	ASSERT(rrl->rr_writer == NULL);
+	refcount_destroy(&rrl->rr_anon_rcount);
+	refcount_destroy(&rrl->rr_linked_rcount);
+}
+
+static void
+rrw_enter_read(rrwlock_t *rrl, void *tag)
+{
+	mutex_enter(&rrl->rr_lock);
+	ASSERT(rrl->rr_writer != curthread);
+	ASSERT(refcount_count(&rrl->rr_anon_rcount) >= 0);
+
+	while (rrl->rr_writer || (rrl->rr_writer_wanted &&
+	    refcount_is_zero(&rrl->rr_anon_rcount) &&
+	    rrn_find(rrl) == NULL))
+		cv_wait(&rrl->rr_cv, &rrl->rr_lock);
+
+	if (rrl->rr_writer_wanted) {
+		/* may or may not be a re-entrant enter */
+		rrn_add(rrl);
+		(void) refcount_add(&rrl->rr_linked_rcount, tag);
+	} else {
+		(void) refcount_add(&rrl->rr_anon_rcount, tag);
+	}
+	ASSERT(rrl->rr_writer == NULL);
+	mutex_exit(&rrl->rr_lock);
+}
+
+static void
+rrw_enter_write(rrwlock_t *rrl)
+{
+	mutex_enter(&rrl->rr_lock);
+	ASSERT(rrl->rr_writer != curthread);
+
+	while (refcount_count(&rrl->rr_anon_rcount) > 0 ||
+	    refcount_count(&rrl->rr_linked_rcount) > 0 ||
+	    rrl->rr_writer != NULL) {
+		rrl->rr_writer_wanted = B_TRUE;
+		cv_wait(&rrl->rr_cv, &rrl->rr_lock);
+	}
+	rrl->rr_writer_wanted = B_FALSE;
+	rrl->rr_writer = curthread;
+	mutex_exit(&rrl->rr_lock);
+}
+
+void
+rrw_enter(rrwlock_t *rrl, krw_t rw, void *tag)
+{
+	if (rw == RW_READER)
+		rrw_enter_read(rrl, tag);
+	else
+		rrw_enter_write(rrl);
+}
+
+void
+rrw_exit(rrwlock_t *rrl, void *tag)
+{
+	mutex_enter(&rrl->rr_lock);
+	ASSERT(!refcount_is_zero(&rrl->rr_anon_rcount) ||
+	    !refcount_is_zero(&rrl->rr_linked_rcount) ||
+	    rrl->rr_writer != NULL);
+
+	if (rrl->rr_writer == NULL) {
+		if (rrn_find_and_remove(rrl)) {
+			if (refcount_remove(&rrl->rr_linked_rcount, tag) == 0)
+				cv_broadcast(&rrl->rr_cv);
+
+		} else {
+			if (refcount_remove(&rrl->rr_anon_rcount, tag) == 0)
+				cv_broadcast(&rrl->rr_cv);
+		}
+	} else {
+		ASSERT(rrl->rr_writer == curthread);
+		ASSERT(refcount_is_zero(&rrl->rr_anon_rcount) &&
+		    refcount_is_zero(&rrl->rr_linked_rcount));
+		rrl->rr_writer = NULL;
+		cv_broadcast(&rrl->rr_cv);
+	}
+	mutex_exit(&rrl->rr_lock);
+}
+
+boolean_t
+rrw_held(rrwlock_t *rrl, krw_t rw)
+{
+	boolean_t held;
+
+	mutex_enter(&rrl->rr_lock);
+	if (rw == RW_WRITER) {
+		held = (rrl->rr_writer == curthread);
+	} else {
+		held = (!refcount_is_zero(&rrl->rr_anon_rcount) ||
+		    !refcount_is_zero(&rrl->rr_linked_rcount));
+	}
+	mutex_exit(&rrl->rr_lock);
+
+	return (held);
+}
--- a/zfs/lib/libdmu-ctl/zfs_acl.c
+++ b/zfs/lib/libdmu-ctl/zfs_acl.c
--- a/zfs/lib/libdmu-ctl/zfs_ctldir.c
+++ b/zfs/lib/libdmu-ctl/zfs_ctldir.c
--- a/zfs/lib/libdmu-ctl/zfs_dir.c
+++ b/zfs/lib/libdmu-ctl/zfs_dir.c
@ -0,0 +1,968 @@
+/*
+ * CDDL HEADER START
+ *
+ * The contents of this file are subject to the terms of the
+ * Common Development and Distribution License (the "License").
+ * You may not use this file except in compliance with the License.
+ *
+ * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
+ * or http://www.opensolaris.org/os/licensing.
+ * See the License for the specific language governing permissions
+ * and limitations under the License.
+ *
+ * When distributing Covered Code, include this CDDL HEADER in each
+ * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
+ * If applicable, add the following below this CDDL HEADER, with the
+ * fields enclosed by brackets "[]" replaced with your own identifying
+ * information: Portions Copyright [yyyy] [name of copyright owner]
+ *
+ * CDDL HEADER END
+ */
+/*
+ * Copyright 2008 Sun Microsystems, Inc.  All rights reserved.
+ * Use is subject to license terms.
+ */
+
+#pragma ident	"@(#)zfs_dir.c	1.25	08/04/27 SMI"
+
+#include <sys/types.h>
+#include <sys/param.h>
+#include <sys/time.h>
+#include <sys/systm.h>
+#include <sys/sysmacros.h>
+#include <sys/resource.h>
+#include <sys/vfs.h>
+#include <sys/vnode.h>
+#include <sys/file.h>
+#include <sys/mode.h>
+#include <sys/kmem.h>
+#include <sys/uio.h>
+#include <sys/pathname.h>
+#include <sys/cmn_err.h>
+#include <sys/errno.h>
+#include <sys/stat.h>
+#include <sys/unistd.h>
+#include <sys/sunddi.h>
+#include <sys/random.h>
+#include <sys/policy.h>
+#include <sys/zfs_dir.h>
+#include <sys/zfs_acl.h>
+#include <sys/fs/zfs.h>
+#include "fs/fs_subr.h"
+#include <sys/zap.h>
+#include <sys/dmu.h>
+#include <sys/atomic.h>
+#include <sys/zfs_ctldir.h>
+#include <sys/zfs_fuid.h>
+#include <sys/dnlc.h>
+#include <sys/extdirent.h>
+
+/*
+ * zfs_match_find() is used by zfs_dirent_lock() to peform zap lookups
+ * of names after deciding which is the appropriate lookup interface.
+ */
+static int
+zfs_match_find(zfsvfs_t *zfsvfs, znode_t *dzp, char *name, boolean_t exact,
+    boolean_t update, int *deflags, pathname_t *rpnp, uint64_t *zoid)
+{
+	int error;
+
+	if (zfsvfs->z_norm) {
+		matchtype_t mt = MT_FIRST;
+		boolean_t conflict = B_FALSE;
+		size_t bufsz = 0;
+		char *buf = NULL;
+
+		if (rpnp) {
+			buf = rpnp->pn_buf;
+			bufsz = rpnp->pn_bufsize;
+		}
+		if (exact)
+			mt = MT_EXACT;
+		/*
+		 * In the non-mixed case we only expect there would ever
+		 * be one match, but we need to use the normalizing lookup.
+		 */
+		error = zap_lookup_norm(zfsvfs->z_os, dzp->z_id, name, 8, 1,
+		    zoid, mt, buf, bufsz, &conflict);
+		if (!error && deflags)
+			*deflags = conflict ? ED_CASE_CONFLICT : 0;
+	} else {
+		error = zap_lookup(zfsvfs->z_os, dzp->z_id, name, 8, 1, zoid);
+	}
+	*zoid = ZFS_DIRENT_OBJ(*zoid);
+
+	if (error == ENOENT && update)
+		dnlc_update(ZTOV(dzp), name, DNLC_NO_VNODE);
+
+	return (error);
+}
+
+/*
+ * Lock a directory entry.  A dirlock on <dzp, name> protects that name
+ * in dzp's directory zap object.  As long as you hold a dirlock, you can
+ * assume two things: (1) dzp cannot be reaped, and (2) no other thread
+ * can change the zap entry for (i.e. link or unlink) this name.
+ *
+ * Input arguments:
+ *	dzp	- znode for directory
+ *	name	- name of entry to lock
+ *	flag	- ZNEW: if the entry already exists, fail with EEXIST.
+ *		  ZEXISTS: if the entry does not exist, fail with ENOENT.
+ *		  ZSHARED: allow concurrent access with other ZSHARED callers.
+ *		  ZXATTR: we want dzp's xattr directory
+ *		  ZCILOOK: On a mixed sensitivity file system,
+ *			   this lookup should be case-insensitive.
+ *		  ZCIEXACT: On a purely case-insensitive file system,
+ *			    this lookup should be case-sensitive.
+ *		  ZRENAMING: we are locking for renaming, force narrow locks
+ *
+ * Output arguments:
+ *	zpp	- pointer to the znode for the entry (NULL if there isn't one)
+ *	dlpp	- pointer to the dirlock for this entry (NULL on error)
+ *      direntflags - (case-insensitive lookup only)
+ *		flags if multiple case-sensitive matches exist in directory
+ *      realpnp     - (case-insensitive lookup only)
+ *		actual name matched within the directory
+ *
+ * Return value: 0 on success or errno on failure.
+ *
+ * NOTE: Always checks for, and rejects, '.' and '..'.
+ * NOTE: For case-insensitive file systems we take wide locks (see below),
+ *	 but return znode pointers to a single match.
+ */
+int
+zfs_dirent_lock(zfs_dirlock_t **dlpp, znode_t *dzp, char *name, znode_t **zpp,
+    int flag, int *direntflags, pathname_t *realpnp)
+{
+	zfsvfs_t	*zfsvfs = dzp->z_zfsvfs;
+	zfs_dirlock_t	*dl;
+	boolean_t	update;
+	boolean_t	exact;
+	uint64_t	zoid;
+	vnode_t		*vp = NULL;
+	int		error = 0;
+	int		cmpflags;
+
+	*zpp = NULL;
+	*dlpp = NULL;
+
+	/*
+	 * Verify that we are not trying to lock '.', '..', or '.zfs'
+	 */
+	if (name[0] == '.' &&
+	    (name[1] == '\0' || (name[1] == '.' && name[2] == '\0')) ||
+	    zfs_has_ctldir(dzp) && strcmp(name, ZFS_CTLDIR_NAME) == 0)
+		return (EEXIST);
+
+	/*
+	 * Case sensitivity and normalization preferences are set when
+	 * the file system is created.  These are stored in the
+	 * zfsvfs->z_case and zfsvfs->z_norm fields.  These choices
+	 * affect what vnodes can be cached in the DNLC, how we
+	 * perform zap lookups, and the "width" of our dirlocks.
+	 *
+	 * A normal dirlock locks a single name.  Note that with
+	 * normalization a name can be composed multiple ways, but
+	 * when normalized, these names all compare equal.  A wide
+	 * dirlock locks multiple names.  We need these when the file
+	 * system is supporting mixed-mode access.  It is sometimes
+	 * necessary to lock all case permutations of file name at
+	 * once so that simultaneous case-insensitive/case-sensitive
+	 * behaves as rationally as possible.
+	 */
+
+	/*
+	 * Decide if exact matches should be requested when performing
+	 * a zap lookup on file systems supporting case-insensitive
+	 * access.
+	 */
+	exact =
+	    ((zfsvfs->z_case == ZFS_CASE_INSENSITIVE) && (flag & ZCIEXACT)) ||
+	    ((zfsvfs->z_case == ZFS_CASE_MIXED) && !(flag & ZCILOOK));
+
+	/*
+	 * Only look in or update the DNLC if we are looking for the
+	 * name on a file system that does not require normalization
+	 * or case folding.  We can also look there if we happen to be
+	 * on a non-normalizing, mixed sensitivity file system IF we
+	 * are looking for the exact name.
+	 *
+	 * Maybe can add TO-UPPERed version of name to dnlc in ci-only
+	 * case for performance improvement?
+	 */
+	update = !zfsvfs->z_norm ||
+	    ((zfsvfs->z_case == ZFS_CASE_MIXED) &&
+	    !(zfsvfs->z_norm & ~U8_TEXTPREP_TOUPPER) && !(flag & ZCILOOK));
+
+	/*
+	 * ZRENAMING indicates we are in a situation where we should
+	 * take narrow locks regardless of the file system's
+	 * preferences for normalizing and case folding.  This will
+	 * prevent us deadlocking trying to grab the same wide lock
+	 * twice if the two names happen to be case-insensitive
+	 * matches.
+	 */
+	if (flag & ZRENAMING)
+		cmpflags = 0;
+	else
+		cmpflags = zfsvfs->z_norm;
+
+	/*
+	 * Wait until there are no locks on this name.
+	 */
+	rw_enter(&dzp->z_name_lock, RW_READER);
+	mutex_enter(&dzp->z_lock);
+	for (;;) {
+		if (dzp->z_unlinked) {
+			mutex_exit(&dzp->z_lock);
+			rw_exit(&dzp->z_name_lock);
+			return (ENOENT);
+		}
+		for (dl = dzp->z_dirlocks; dl != NULL; dl = dl->dl_next) {
+			if ((u8_strcmp(name, dl->dl_name, 0, cmpflags,
+			    U8_UNICODE_LATEST, &error) == 0) || error != 0)
+				break;
+		}
+		if (error != 0) {
+			mutex_exit(&dzp->z_lock);
+			rw_exit(&dzp->z_name_lock);
+			return (ENOENT);
+		}
+		if (dl == NULL)	{
+			/*
+			 * Allocate a new dirlock and add it to the list.
+			 */
+			dl = kmem_alloc(sizeof (zfs_dirlock_t), KM_SLEEP);
+			cv_init(&dl->dl_cv, NULL, CV_DEFAULT, NULL);
+			dl->dl_name = name;
+			dl->dl_sharecnt = 0;
+			dl->dl_namesize = 0;
+			dl->dl_dzp = dzp;
+			dl->dl_next = dzp->z_dirlocks;
+			dzp->z_dirlocks = dl;
+			break;
+		}
+		if ((flag & ZSHARED) && dl->dl_sharecnt != 0)
+			break;
+		cv_wait(&dl->dl_cv, &dzp->z_lock);
+	}
+
+	if ((flag & ZSHARED) && ++dl->dl_sharecnt > 1 && dl->dl_namesize == 0) {
+		/*
+		 * We're the second shared reference to dl.  Make a copy of
+		 * dl_name in case the first thread goes away before we do.
+		 * Note that we initialize the new name before storing its
+		 * pointer into dl_name, because the first thread may load
+		 * dl->dl_name at any time.  He'll either see the old value,
+		 * which is his, or the new shared copy; either is OK.
+		 */
+		dl->dl_namesize = strlen(dl->dl_name) + 1;
+		name = kmem_alloc(dl->dl_namesize, KM_SLEEP);
+		bcopy(dl->dl_name, name, dl->dl_namesize);
+		dl->dl_name = name;
+	}
+
+	mutex_exit(&dzp->z_lock);
+
+	/*
+	 * We have a dirlock on the name.  (Note that it is the dirlock,
+	 * not the dzp's z_lock, that protects the name in the zap object.)
+	 * See if there's an object by this name; if so, put a hold on it.
+	 */
+	if (flag & ZXATTR) {
+		zoid = dzp->z_phys->zp_xattr;
+		error = (zoid == 0 ? ENOENT : 0);
+	} else {
+		if (update)
+			vp = dnlc_lookup(ZTOV(dzp), name);
+		if (vp == DNLC_NO_VNODE) {
+			VN_RELE(vp);
+			error = ENOENT;
+		} else if (vp) {
+			if (flag & ZNEW) {
+				zfs_dirent_unlock(dl);
+				VN_RELE(vp);
+				return (EEXIST);
+			}
+			*dlpp = dl;
+			*zpp = VTOZ(vp);
+			return (0);
+		} else {
+			error = zfs_match_find(zfsvfs, dzp, name, exact,
+			    update, direntflags, realpnp, &zoid);
+		}
+	}
+	if (error) {
+		if (error != ENOENT || (flag & ZEXISTS)) {
+			zfs_dirent_unlock(dl);
+			return (error);
+		}
+	} else {
+		if (flag & ZNEW) {
+			zfs_dirent_unlock(dl);
+			return (EEXIST);
+		}
+		error = zfs_zget(zfsvfs, zoid, zpp);
+		if (error) {
+			zfs_dirent_unlock(dl);
+			return (error);
+		}
+		if (!(flag & ZXATTR) && update)
+			dnlc_update(ZTOV(dzp), name, ZTOV(*zpp));
+	}
+
+	*dlpp = dl;
+
+	return (0);
+}
+
+/*
+ * Unlock this directory entry and wake anyone who was waiting for it.
+ */
+void
+zfs_dirent_unlock(zfs_dirlock_t *dl)
+{
+	znode_t *dzp = dl->dl_dzp;
+	zfs_dirlock_t **prev_dl, *cur_dl;
+
+	mutex_enter(&dzp->z_lock);
+	rw_exit(&dzp->z_name_lock);
+	if (dl->dl_sharecnt > 1) {
+		dl->dl_sharecnt--;
+		mutex_exit(&dzp->z_lock);
+		return;
+	}
+	prev_dl = &dzp->z_dirlocks;
+	while ((cur_dl = *prev_dl) != dl)
+		prev_dl = &cur_dl->dl_next;
+	*prev_dl = dl->dl_next;
+	cv_broadcast(&dl->dl_cv);
+	mutex_exit(&dzp->z_lock);
+
+	if (dl->dl_namesize != 0)
+		kmem_free(dl->dl_name, dl->dl_namesize);
+	cv_destroy(&dl->dl_cv);
+	kmem_free(dl, sizeof (*dl));
+}
+
+/*
+ * Look up an entry in a directory.
+ *
+ * NOTE: '.' and '..' are handled as special cases because
+ *	no directory entries are actually stored for them.  If this is
+ *	the root of a filesystem, then '.zfs' is also treated as a
+ *	special pseudo-directory.
+ */
+int
+zfs_dirlook(znode_t *dzp, char *name, vnode_t **vpp, int flags,
+    int *deflg, pathname_t *rpnp)
+{
+	zfs_dirlock_t *dl;
+	znode_t *zp;
+	int error = 0;
+
+	if (name[0] == 0 || (name[0] == '.' && name[1] == 0)) {
+		*vpp = ZTOV(dzp);
+		VN_HOLD(*vpp);
+	} else if (name[0] == '.' && name[1] == '.' && name[2] == 0) {
+		zfsvfs_t *zfsvfs = dzp->z_zfsvfs;
+		/*
+		 * If we are a snapshot mounted under .zfs, return
+		 * the vp for the snapshot directory.
+		 */
+		if (dzp->z_phys->zp_parent == dzp->z_id &&
+		    zfsvfs->z_parent != zfsvfs) {
+			error = zfsctl_root_lookup(zfsvfs->z_parent->z_ctldir,
+			    "snapshot", vpp, NULL, 0, NULL, kcred,
+			    NULL, NULL, NULL);
+			return (error);
+		}
+		rw_enter(&dzp->z_parent_lock, RW_READER);
+		error = zfs_zget(zfsvfs, dzp->z_phys->zp_parent, &zp);
+		if (error == 0)
+			*vpp = ZTOV(zp);
+		rw_exit(&dzp->z_parent_lock);
+	} else if (zfs_has_ctldir(dzp) && strcmp(name, ZFS_CTLDIR_NAME) == 0) {
+		*vpp = zfsctl_root(dzp);
+	} else {
+		int zf;
+
+		zf = ZEXISTS | ZSHARED;
+		if (flags & FIGNORECASE)
+			zf |= ZCILOOK;
+
+		error = zfs_dirent_lock(&dl, dzp, name, &zp, zf, deflg, rpnp);
+		if (error == 0) {
+			*vpp = ZTOV(zp);
+			zfs_dirent_unlock(dl);
+			dzp->z_zn_prefetch = B_TRUE; /* enable prefetching */
+		}
+		rpnp = NULL;
+	}
+
+	if ((flags & FIGNORECASE) && rpnp && !error)
+		(void) strlcpy(rpnp->pn_buf, name, rpnp->pn_bufsize);
+
+	return (error);
+}
+
+static char *
+zfs_unlinked_hexname(char namebuf[17], uint64_t x)
+{
+	char *name = &namebuf[16];
+	const char digits[16] = "0123456789abcdef";
+
+	*name = '\0';
+	do {
+		*--name = digits[x & 0xf];
+		x >>= 4;
+	} while (x != 0);
+
+	return (name);
+}
+
+/*
+ * unlinked Set (formerly known as the "delete queue") Error Handling
+ *
+ * When dealing with the unlinked set, we dmu_tx_hold_zap(), but we
+ * don't specify the name of the entry that we will be manipulating.  We
+ * also fib and say that we won't be adding any new entries to the
+ * unlinked set, even though we might (this is to lower the minimum file
+ * size that can be deleted in a full filesystem).  So on the small
+ * chance that the nlink list is using a fat zap (ie. has more than
+ * 2000 entries), we *may* not pre-read a block that's needed.
+ * Therefore it is remotely possible for some of the assertions
+ * regarding the unlinked set below to fail due to i/o error.  On a
+ * nondebug system, this will result in the space being leaked.
+ */
+void
+zfs_unlinked_add(znode_t *zp, dmu_tx_t *tx)
+{
+	zfsvfs_t *zfsvfs = zp->z_zfsvfs;
+	char obj_name[17];
+	int error;
+
+	ASSERT(zp->z_unlinked);
+	ASSERT3U(zp->z_phys->zp_links, ==, 0);
+
+	error = zap_add(zfsvfs->z_os, zfsvfs->z_unlinkedobj,
+	    zfs_unlinked_hexname(obj_name, zp->z_id), 8, 1, &zp->z_id, tx);
+	ASSERT3U(error, ==, 0);
+}
+
+/*
+ * Clean up any znodes that had no links when we either crashed or
+ * (force) umounted the file system.
+ */
+void
+zfs_unlinked_drain(zfsvfs_t *zfsvfs)
+{
+	zap_cursor_t	zc;
+	zap_attribute_t zap;
+	dmu_object_info_t doi;
+	znode_t		*zp;
+	int		error;
+
+	/*
+	 * Interate over the contents of the unlinked set.
+	 */
+	for (zap_cursor_init(&zc, zfsvfs->z_os, zfsvfs->z_unlinkedobj);
+	    zap_cursor_retrieve(&zc, &zap) == 0;
+	    zap_cursor_advance(&zc)) {
+
+		/*
+		 * See what kind of object we have in list
+		 */
+
+		error = dmu_object_info(zfsvfs->z_os,
+		    zap.za_first_integer, &doi);
+		if (error != 0)
+			continue;
+
+		ASSERT((doi.doi_type == DMU_OT_PLAIN_FILE_CONTENTS) ||
+		    (doi.doi_type == DMU_OT_DIRECTORY_CONTENTS));
+		/*
+		 * We need to re-mark these list entries for deletion,
+		 * so we pull them back into core and set zp->z_unlinked.
+		 */
+		error = zfs_zget(zfsvfs, zap.za_first_integer, &zp);
+
+		/*
+		 * We may pick up znodes that are already marked for deletion.
+		 * This could happen during the purge of an extended attribute
+		 * directory.  All we need to do is skip over them, since they
+		 * are already in the system marked z_unlinked.
+		 */
+		if (error != 0)
+			continue;
+
+		zp->z_unlinked = B_TRUE;
+		VN_RELE(ZTOV(zp));
+	}
+	zap_cursor_fini(&zc);
+}
+
+/*
+ * Delete the entire contents of a directory.  Return a count
+ * of the number of entries that could not be deleted. If we encounter
+ * an error, return a count of at least one so that the directory stays
+ * in the unlinked set.
+ *
+ * NOTE: this function assumes that the directory is inactive,
+ *	so there is no need to lock its entries before deletion.
+ *	Also, it assumes the directory contents is *only* regular
+ *	files.
+ */
+static int
+zfs_purgedir(znode_t *dzp)
+{
+	zap_cursor_t	zc;
+	zap_attribute_t	zap;
+	znode_t		*xzp;
+	dmu_tx_t	*tx;
+	zfsvfs_t	*zfsvfs = dzp->z_zfsvfs;
+	zfs_dirlock_t	dl;
+	int skipped = 0;
+	int error;
+
+	for (zap_cursor_init(&zc, zfsvfs->z_os, dzp->z_id);
+	    (error = zap_cursor_retrieve(&zc, &zap)) == 0;
+	    zap_cursor_advance(&zc)) {
+		error = zfs_zget(zfsvfs,
+		    ZFS_DIRENT_OBJ(zap.za_first_integer), &xzp);
+		if (error) {
+			skipped += 1;
+			continue;
+		}
+
+		ASSERT((ZTOV(xzp)->v_type == VREG) ||
+		    (ZTOV(xzp)->v_type == VLNK));
+
+		tx = dmu_tx_create(zfsvfs->z_os);
+		dmu_tx_hold_bonus(tx, dzp->z_id);
+		dmu_tx_hold_zap(tx, dzp->z_id, FALSE, zap.za_name);
+		dmu_tx_hold_bonus(tx, xzp->z_id);
+		dmu_tx_hold_zap(tx, zfsvfs->z_unlinkedobj, FALSE, NULL);
+		error = dmu_tx_assign(tx, TXG_WAIT);
+		if (error) {
+			dmu_tx_abort(tx);
+			VN_RELE(ZTOV(xzp));
+			skipped += 1;
+			continue;
+		}
+		bzero(&dl, sizeof (dl));
+		dl.dl_dzp = dzp;
+		dl.dl_name = zap.za_name;
+
+		error = zfs_link_destroy(&dl, xzp, tx, 0, NULL);
+		if (error)
+			skipped += 1;
+		dmu_tx_commit(tx);
+
+		VN_RELE(ZTOV(xzp));
+	}
+	zap_cursor_fini(&zc);
+	if (error != ENOENT)
+		skipped += 1;
+	return (skipped);
+}
+
+void
+zfs_rmnode(znode_t *zp)
+{
+	zfsvfs_t	*zfsvfs = zp->z_zfsvfs;
+	objset_t	*os = zfsvfs->z_os;
+	znode_t		*xzp = NULL;
+	char		obj_name[17];
+	dmu_tx_t	*tx;
+	uint64_t	acl_obj;
+	int		error;
+
+	ASSERT(ZTOV(zp)->v_count == 0);
+	ASSERT(zp->z_phys->zp_links == 0);
+
+	/*
+	 * If this is an attribute directory, purge its contents.
+	 */
+	if (ZTOV(zp)->v_type == VDIR && (zp->z_phys->zp_flags & ZFS_XATTR)) {
+		if (zfs_purgedir(zp) != 0) {
+			/*
+			 * Not enough space to delete some xattrs.
+			 * Leave it on the unlinked set.
+			 */
+			zfs_znode_dmu_fini(zp);
+			zfs_znode_free(zp);
+			return;
+		}
+	}
+
+	/*
+	 * If the file has extended attributes, we're going to unlink
+	 * the xattr dir.
+	 */
+	if (zp->z_phys->zp_xattr) {
+		error = zfs_zget(zfsvfs, zp->z_phys->zp_xattr, &xzp);
+		ASSERT(error == 0);
+	}
+
+	acl_obj = zp->z_phys->zp_acl.z_acl_extern_obj;
+
+	/*
+	 * Set up the transaction.
+	 */
+	tx = dmu_tx_create(os);
+	dmu_tx_hold_free(tx, zp->z_id, 0, DMU_OBJECT_END);
+	dmu_tx_hold_zap(tx, zfsvfs->z_unlinkedobj, FALSE, NULL);
+	if (xzp) {
+		dmu_tx_hold_bonus(tx, xzp->z_id);
+		dmu_tx_hold_zap(tx, zfsvfs->z_unlinkedobj, TRUE, NULL);
+	}
+	if (acl_obj)
+		dmu_tx_hold_free(tx, acl_obj, 0, DMU_OBJECT_END);
+	error = dmu_tx_assign(tx, TXG_WAIT);
+	if (error) {
+		/*
+		 * Not enough space to delete the file.  Leave it in the
+		 * unlinked set, leaking it until the fs is remounted (at
+		 * which point we'll call zfs_unlinked_drain() to process it).
+		 */
+		dmu_tx_abort(tx);
+		zfs_znode_dmu_fini(zp);
+		zfs_znode_free(zp);
+		goto out;
+	}
+
+	if (xzp) {
+		dmu_buf_will_dirty(xzp->z_dbuf, tx);
+		mutex_enter(&xzp->z_lock);
+		xzp->z_unlinked = B_TRUE;	/* mark xzp for deletion */
+		xzp->z_phys->zp_links = 0;	/* no more links to it */
+		mutex_exit(&xzp->z_lock);
+		zfs_unlinked_add(xzp, tx);
+	}
+
+	/* Remove this znode from the unlinked set */
+	error = zap_remove(os, zfsvfs->z_unlinkedobj,
+	    zfs_unlinked_hexname(obj_name, zp->z_id), tx);
+	ASSERT3U(error, ==, 0);
+
+	zfs_znode_delete(zp, tx);
+
+	dmu_tx_commit(tx);
+out:
+	if (xzp)
+		VN_RELE(ZTOV(xzp));
+}
+
+static uint64_t
+zfs_dirent(znode_t *zp)
+{
+	uint64_t de = zp->z_id;
+	if (zp->z_zfsvfs->z_version >= ZPL_VERSION_DIRENT_TYPE)
+		de |= IFTODT((zp)->z_phys->zp_mode) << 60;
+	return (de);
+}
+
+/*
+ * Link zp into dl.  Can only fail if zp has been unlinked.
+ */
+int
+zfs_link_create(zfs_dirlock_t *dl, znode_t *zp, dmu_tx_t *tx, int flag)
+{
+	znode_t *dzp = dl->dl_dzp;
+	vnode_t *vp = ZTOV(zp);
+	uint64_t value;
+	int zp_is_dir = (vp->v_type == VDIR);
+	int error;
+
+	dmu_buf_will_dirty(zp->z_dbuf, tx);
+	mutex_enter(&zp->z_lock);
+
+	if (!(flag & ZRENAMING)) {
+		if (zp->z_unlinked) {	/* no new links to unlinked zp */
+			ASSERT(!(flag & (ZNEW | ZEXISTS)));
+			mutex_exit(&zp->z_lock);
+			return (ENOENT);
+		}
+		zp->z_phys->zp_links++;
+	}
+	zp->z_phys->zp_parent = dzp->z_id;	/* dzp is now zp's parent */
+
+	if (!(flag & ZNEW))
+		zfs_time_stamper_locked(zp, STATE_CHANGED, tx);
+	mutex_exit(&zp->z_lock);
+
+	dmu_buf_will_dirty(dzp->z_dbuf, tx);
+	mutex_enter(&dzp->z_lock);
+	dzp->z_phys->zp_size++;			/* one dirent added */
+	dzp->z_phys->zp_links += zp_is_dir;	/* ".." link from zp */
+	zfs_time_stamper_locked(dzp, CONTENT_MODIFIED, tx);
+	mutex_exit(&dzp->z_lock);
+
+	value = zfs_dirent(zp);
+	error = zap_add(zp->z_zfsvfs->z_os, dzp->z_id, dl->dl_name,
+	    8, 1, &value, tx);
+	ASSERT(error == 0);
+
+	dnlc_update(ZTOV(dzp), dl->dl_name, vp);
+
+	return (0);
+}
+
+/*
+ * Unlink zp from dl, and mark zp for deletion if this was the last link.
+ * Can fail if zp is a mount point (EBUSY) or a non-empty directory (EEXIST).
+ * If 'unlinkedp' is NULL, we put unlinked znodes on the unlinked list.
+ * If it's non-NULL, we use it to indicate whether the znode needs deletion,
+ * and it's the caller's job to do it.
+ */
+int
+zfs_link_destroy(zfs_dirlock_t *dl, znode_t *zp, dmu_tx_t *tx, int flag,
+	boolean_t *unlinkedp)
+{
+	znode_t *dzp = dl->dl_dzp;
+	vnode_t *vp = ZTOV(zp);
+	int zp_is_dir = (vp->v_type == VDIR);
+	boolean_t unlinked = B_FALSE;
+	int error;
+
+	dnlc_remove(ZTOV(dzp), dl->dl_name);
+
+	if (!(flag & ZRENAMING)) {
+		dmu_buf_will_dirty(zp->z_dbuf, tx);
+
+		if (vn_vfswlock(vp))		/* prevent new mounts on zp */
+			return (EBUSY);
+
+		if (vn_ismntpt(vp)) {		/* don't remove mount point */
+			vn_vfsunlock(vp);
+			return (EBUSY);
+		}
+
+		mutex_enter(&zp->z_lock);
+		if (zp_is_dir && !zfs_dirempty(zp)) {	/* dir not empty */
+			mutex_exit(&zp->z_lock);
+			vn_vfsunlock(vp);
+			return (EEXIST);
+		}
+		if (zp->z_phys->zp_links <= zp_is_dir) {
+			zfs_panic_recover("zfs: link count on %s is %u, "
+			    "should be at least %u",
+			    zp->z_vnode->v_path ? zp->z_vnode->v_path :
+			    "<unknown>", (int)zp->z_phys->zp_links,
+			    zp_is_dir + 1);
+			zp->z_phys->zp_links = zp_is_dir + 1;
+		}
+		if (--zp->z_phys->zp_links == zp_is_dir) {
+			zp->z_unlinked = B_TRUE;
+			zp->z_phys->zp_links = 0;
+			unlinked = B_TRUE;
+		} else {
+			zfs_time_stamper_locked(zp, STATE_CHANGED, tx);
+		}
+		mutex_exit(&zp->z_lock);
+		vn_vfsunlock(vp);
+	}
+
+	dmu_buf_will_dirty(dzp->z_dbuf, tx);
+	mutex_enter(&dzp->z_lock);
+	dzp->z_phys->zp_size--;			/* one dirent removed */
+	dzp->z_phys->zp_links -= zp_is_dir;	/* ".." link from zp */
+	zfs_time_stamper_locked(dzp, CONTENT_MODIFIED, tx);
+	mutex_exit(&dzp->z_lock);
+
+	if (zp->z_zfsvfs->z_norm) {
+		if (((zp->z_zfsvfs->z_case == ZFS_CASE_INSENSITIVE) &&
+		    (flag & ZCIEXACT)) ||
+		    ((zp->z_zfsvfs->z_case == ZFS_CASE_MIXED) &&
+		    !(flag & ZCILOOK)))
+			error = zap_remove_norm(zp->z_zfsvfs->z_os,
+			    dzp->z_id, dl->dl_name, MT_EXACT, tx);
+		else
+			error = zap_remove_norm(zp->z_zfsvfs->z_os,
+			    dzp->z_id, dl->dl_name, MT_FIRST, tx);
+	} else {
+		error = zap_remove(zp->z_zfsvfs->z_os,
+		    dzp->z_id, dl->dl_name, tx);
+	}
+	ASSERT(error == 0);
+
+	if (unlinkedp != NULL)
+		*unlinkedp = unlinked;
+	else if (unlinked)
+		zfs_unlinked_add(zp, tx);
+
+	return (0);
+}
+
+/*
+ * Indicate whether the directory is empty.  Works with or without z_lock
+ * held, but can only be consider a hint in the latter case.  Returns true
+ * if only "." and ".." remain and there's no work in progress.
+ */
+boolean_t
+zfs_dirempty(znode_t *dzp)
+{
+	return (dzp->z_phys->zp_size == 2 && dzp->z_dirlocks == 0);
+}
+
+int
+zfs_make_xattrdir(znode_t *zp, vattr_t *vap, vnode_t **xvpp, cred_t *cr)
+{
+	zfsvfs_t *zfsvfs = zp->z_zfsvfs;
+	znode_t *xzp;
+	dmu_tx_t *tx;
+	int error;
+	zfs_fuid_info_t *fuidp = NULL;
+
+	*xvpp = NULL;
+
+	if (error = zfs_zaccess(zp, ACE_WRITE_NAMED_ATTRS, 0, B_FALSE, cr))
+		return (error);
+
+	tx = dmu_tx_create(zfsvfs->z_os);
+	dmu_tx_hold_bonus(tx, zp->z_id);
+	dmu_tx_hold_zap(tx, DMU_NEW_OBJECT, FALSE, NULL);
+	if (IS_EPHEMERAL(crgetuid(cr)) || IS_EPHEMERAL(crgetgid(cr))) {
+		if (zfsvfs->z_fuid_obj == 0) {
+			dmu_tx_hold_bonus(tx, DMU_NEW_OBJECT);
+			dmu_tx_hold_write(tx, DMU_NEW_OBJECT, 0,
+			    FUID_SIZE_ESTIMATE(zfsvfs));
+			dmu_tx_hold_zap(tx, MASTER_NODE_OBJ, FALSE, NULL);
+		} else {
+			dmu_tx_hold_bonus(tx, zfsvfs->z_fuid_obj);
+			dmu_tx_hold_write(tx, zfsvfs->z_fuid_obj, 0,
+			    FUID_SIZE_ESTIMATE(zfsvfs));
+		}
+	}
+	error = dmu_tx_assign(tx, zfsvfs->z_assign);
+	if (error) {
+		if (error == ERESTART && zfsvfs->z_assign == TXG_NOWAIT)
+			dmu_tx_wait(tx);
+		dmu_tx_abort(tx);
+		return (error);
+	}
+	zfs_mknode(zp, vap, tx, cr, IS_XATTR, &xzp, 0, NULL, &fuidp);
+	ASSERT(xzp->z_phys->zp_parent == zp->z_id);
+	dmu_buf_will_dirty(zp->z_dbuf, tx);
+	zp->z_phys->zp_xattr = xzp->z_id;
+
+	(void) zfs_log_create(zfsvfs->z_log, tx, TX_MKXATTR, zp,
+	    xzp, "", NULL, fuidp, vap);
+	if (fuidp)
+		zfs_fuid_info_free(fuidp);
+	dmu_tx_commit(tx);
+
+	*xvpp = ZTOV(xzp);
+
+	return (0);
+}
+
+/*
+ * Return a znode for the extended attribute directory for zp.
+ * ** If the directory does not already exist, it is created **
+ *
+ *	IN:	zp	- znode to obtain attribute directory from
+ *		cr	- credentials of caller
+ *		flags	- flags from the VOP_LOOKUP call
+ *
+ *	OUT:	xzpp	- pointer to extended attribute znode
+ *
+ *	RETURN:	0 on success
+ *		error number on failure
+ */
+int
+zfs_get_xattrdir(znode_t *zp, vnode_t **xvpp, cred_t *cr, int flags)
+{
+	zfsvfs_t	*zfsvfs = zp->z_zfsvfs;
+	znode_t		*xzp;
+	zfs_dirlock_t	*dl;
+	vattr_t		va;
+	int		error;
+top:
+	error = zfs_dirent_lock(&dl, zp, "", &xzp, ZXATTR, NULL, NULL);
+	if (error)
+		return (error);
+
+	if (xzp != NULL) {
+		*xvpp = ZTOV(xzp);
+		zfs_dirent_unlock(dl);
+		return (0);
+	}
+
+	ASSERT(zp->z_phys->zp_xattr == 0);
+
+	if (!(flags & CREATE_XATTR_DIR)) {
+		zfs_dirent_unlock(dl);
+		return (ENOENT);
+	}
+
+	if (zfsvfs->z_vfs->vfs_flag & VFS_RDONLY) {
+		zfs_dirent_unlock(dl);
+		return (EROFS);
+	}
+
+	/*
+	 * The ability to 'create' files in an attribute
+	 * directory comes from the write_xattr permission on the base file.
+	 *
+	 * The ability to 'search' an attribute directory requires
+	 * read_xattr permission on the base file.
+	 *
+	 * Once in a directory the ability to read/write attributes
+	 * is controlled by the permissions on the attribute file.
+	 */
+	va.va_mask = AT_TYPE | AT_MODE | AT_UID | AT_GID;
+	va.va_type = VDIR;
+	va.va_mode = S_IFDIR | S_ISVTX | 0777;
+	zfs_fuid_map_ids(zp, cr, &va.va_uid, &va.va_gid);
+
+	error = zfs_make_xattrdir(zp, &va, xvpp, cr);
+	zfs_dirent_unlock(dl);
+
+	if (error == ERESTART && zfsvfs->z_assign == TXG_NOWAIT) {
+		/* NB: we already did dmu_tx_wait() if necessary */
+		goto top;
+	}
+
+	return (error);
+}
+
+/*
+ * Decide whether it is okay to remove within a sticky directory.
+ *
+ * In sticky directories, write access is not sufficient;
+ * you can remove entries from a directory only if:
+ *
+ *	you own the directory,
+ *	you own the entry,
+ *	the entry is a plain file and you have write access,
+ *	or you are privileged (checked in secpolicy...).
+ *
+ * The function returns 0 if remove access is granted.
+ */
+int
+zfs_sticky_remove_access(znode_t *zdp, znode_t *zp, cred_t *cr)
+{
+	uid_t  		uid;
+	uid_t		downer;
+	uid_t		fowner;
+	zfsvfs_t	*zfsvfs = zdp->z_zfsvfs;
+
+	if (zdp->z_zfsvfs->z_assign >= TXG_INITIAL)	/* ZIL replay */
+		return (0);
+
+	if ((zdp->z_phys->zp_mode & S_ISVTX) == 0)
+		return (0);
+
+	downer = zfs_fuid_map_id(zfsvfs, zdp->z_phys->zp_uid, cr, ZFS_OWNER);
+	fowner = zfs_fuid_map_id(zfsvfs, zp->z_phys->zp_uid, cr, ZFS_OWNER);
+
+	if ((uid = crgetuid(cr)) == downer || uid == fowner ||
+	    (ZTOV(zp)->v_type == VREG &&
+	    zfs_zaccess(zp, ACE_WRITE_DATA, 0, B_FALSE, cr) == 0))
+		return (0);
+	else
+		return (secpolicy_vnode_remove(cr));
+}
--- a/zfs/lib/libdmu-ctl/zfs_fuid.c
+++ b/zfs/lib/libdmu-ctl/zfs_fuid.c
@ -0,0 +1,688 @@
+/*
+ * CDDL HEADER START
+ *
+ * The contents of this file are subject to the terms of the
+ * Common Development and Distribution License (the "License").
+ * You may not use this file except in compliance with the License.
+ *
+ * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
+ * or http://www.opensolaris.org/os/licensing.
+ * See the License for the specific language governing permissions
+ * and limitations under the License.
+ *
+ * When distributing Covered Code, include this CDDL HEADER in each
+ * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
+ * If applicable, add the following below this CDDL HEADER, with the
+ * fields enclosed by brackets "[]" replaced with your own identifying
+ * information: Portions Copyright [yyyy] [name of copyright owner]
+ *
+ * CDDL HEADER END
+ */
+/*
+ * Copyright 2008 Sun Microsystems, Inc.  All rights reserved.
+ * Use is subject to license terms.
+ */
+
+#pragma ident	"@(#)zfs_fuid.c	1.5	08/01/31 SMI"
+
+#include <sys/zfs_context.h>
+#include <sys/sunddi.h>
+#include <sys/dmu.h>
+#include <sys/avl.h>
+#include <sys/zap.h>
+#include <sys/refcount.h>
+#include <sys/nvpair.h>
+#ifdef _KERNEL
+#include <sys/kidmap.h>
+#include <sys/sid.h>
+#include <sys/zfs_vfsops.h>
+#include <sys/zfs_znode.h>
+#endif
+#include <sys/zfs_fuid.h>
+
+/*
+ * FUID Domain table(s).
+ *
+ * The FUID table is stored as a packed nvlist of an array
+ * of nvlists which contain an index, domain string and offset
+ *
+ * During file system initialization the nvlist(s) are read and
+ * two AVL trees are created.  One tree is keyed by the index number
+ * and the other by the domain string.  Nodes are never removed from
+ * trees, but new entries may be added.  If a new entry is added then the
+ * on-disk packed nvlist will also be updated.
+ */
+
+#define	FUID_IDX	"fuid_idx"
+#define	FUID_DOMAIN	"fuid_domain"
+#define	FUID_OFFSET	"fuid_offset"
+#define	FUID_NVP_ARRAY	"fuid_nvlist"
+
+typedef struct fuid_domain {
+	avl_node_t	f_domnode;
+	avl_node_t	f_idxnode;
+	ksiddomain_t	*f_ksid;
+	uint64_t	f_idx;
+} fuid_domain_t;
+
+/*
+ * Compare two indexes.
+ */
+static int
+idx_compare(const void *arg1, const void *arg2)
+{
+	const fuid_domain_t *node1 = arg1;
+	const fuid_domain_t *node2 = arg2;
+
+	if (node1->f_idx < node2->f_idx)
+		return (-1);
+	else if (node1->f_idx > node2->f_idx)
+		return (1);
+	return (0);
+}
+
+/*
+ * Compare two domain strings.
+ */
+static int
+domain_compare(const void *arg1, const void *arg2)
+{
+	const fuid_domain_t *node1 = arg1;
+	const fuid_domain_t *node2 = arg2;
+	int val;
+
+	val = strcmp(node1->f_ksid->kd_name, node2->f_ksid->kd_name);
+	if (val == 0)
+		return (0);
+	return (val > 0 ? 1 : -1);
+}
+
+/*
+ * load initial fuid domain and idx trees.  This function is used by
+ * both the kernel and zdb.
+ */
+uint64_t
+zfs_fuid_table_load(objset_t *os, uint64_t fuid_obj, avl_tree_t *idx_tree,
+    avl_tree_t *domain_tree)
+{
+	dmu_buf_t *db;
+	uint64_t fuid_size;
+
+	avl_create(idx_tree, idx_compare,
+	    sizeof (fuid_domain_t), offsetof(fuid_domain_t, f_idxnode));
+	avl_create(domain_tree, domain_compare,
+	    sizeof (fuid_domain_t), offsetof(fuid_domain_t, f_domnode));
+
+	VERIFY(0 == dmu_bonus_hold(os, fuid_obj, FTAG, &db));
+	fuid_size = *(uint64_t *)db->db_data;
+	dmu_buf_rele(db, FTAG);
+
+	if (fuid_size)  {
+		nvlist_t **fuidnvp;
+		nvlist_t *nvp = NULL;
+		uint_t count;
+		char *packed;
+		int i;
+
+		packed = kmem_alloc(fuid_size, KM_SLEEP);
+		VERIFY(dmu_read(os, fuid_obj, 0, fuid_size, packed) == 0);
+		VERIFY(nvlist_unpack(packed, fuid_size,
+		    &nvp, 0) == 0);
+		VERIFY(nvlist_lookup_nvlist_array(nvp, FUID_NVP_ARRAY,
+		    &fuidnvp, &count) == 0);
+
+		for (i = 0; i != count; i++) {
+			fuid_domain_t *domnode;
+			char *domain;
+			uint64_t idx;
+
+			VERIFY(nvlist_lookup_string(fuidnvp[i], FUID_DOMAIN,
+			    &domain) == 0);
+			VERIFY(nvlist_lookup_uint64(fuidnvp[i], FUID_IDX,
+			    &idx) == 0);
+
+			domnode = kmem_alloc(sizeof (fuid_domain_t), KM_SLEEP);
+
+			domnode->f_idx = idx;
+			domnode->f_ksid = ksid_lookupdomain(domain);
+			avl_add(idx_tree, domnode);
+			avl_add(domain_tree, domnode);
+		}
+		nvlist_free(nvp);
+		kmem_free(packed, fuid_size);
+	}
+	return (fuid_size);
+}
+
+void
+zfs_fuid_table_destroy(avl_tree_t *idx_tree, avl_tree_t *domain_tree)
+{
+	fuid_domain_t *domnode;
+	void *cookie;
+
+	cookie = NULL;
+	while (domnode = avl_destroy_nodes(domain_tree, &cookie))
+		ksiddomain_rele(domnode->f_ksid);
+
+	avl_destroy(domain_tree);
+	cookie = NULL;
+	while (domnode = avl_destroy_nodes(idx_tree, &cookie))
+		kmem_free(domnode, sizeof (fuid_domain_t));
+	avl_destroy(idx_tree);
+}
+
+char *
+zfs_fuid_idx_domain(avl_tree_t *idx_tree, uint32_t idx)
+{
+	fuid_domain_t searchnode, *findnode;
+	avl_index_t loc;
+
+	searchnode.f_idx = idx;
+
+	findnode = avl_find(idx_tree, &searchnode, &loc);
+
+	return (findnode->f_ksid->kd_name);
+}
+
+#ifdef _KERNEL
+/*
+ * Load the fuid table(s) into memory.
+ */
+static void
+zfs_fuid_init(zfsvfs_t *zfsvfs, dmu_tx_t *tx)
+{
+	int error = 0;
+
+	rw_enter(&zfsvfs->z_fuid_lock, RW_WRITER);
+
+	if (zfsvfs->z_fuid_loaded) {
+		rw_exit(&zfsvfs->z_fuid_lock);
+		return;
+	}
+
+	if (zfsvfs->z_fuid_obj == 0) {
+
+		/* first make sure we need to allocate object */
+
+		error = zap_lookup(zfsvfs->z_os, MASTER_NODE_OBJ,
+		    ZFS_FUID_TABLES, 8, 1, &zfsvfs->z_fuid_obj);
+		if (error == ENOENT && tx != NULL) {
+			zfsvfs->z_fuid_obj = dmu_object_alloc(zfsvfs->z_os,
+			    DMU_OT_FUID, 1 << 14, DMU_OT_FUID_SIZE,
+			    sizeof (uint64_t), tx);
+			VERIFY(zap_add(zfsvfs->z_os, MASTER_NODE_OBJ,
+			    ZFS_FUID_TABLES, sizeof (uint64_t), 1,
+			    &zfsvfs->z_fuid_obj, tx) == 0);
+		}
+	}
+
+	zfsvfs->z_fuid_size = zfs_fuid_table_load(zfsvfs->z_os,
+	    zfsvfs->z_fuid_obj, &zfsvfs->z_fuid_idx, &zfsvfs->z_fuid_domain);
+
+	zfsvfs->z_fuid_loaded = B_TRUE;
+	rw_exit(&zfsvfs->z_fuid_lock);
+}
+
+/*
+ * Query domain table for a given domain.
+ *
+ * If domain isn't found it is added to AVL trees and
+ * the results are pushed out to disk.
+ */
+int
+zfs_fuid_find_by_domain(zfsvfs_t *zfsvfs, const char *domain, char **retdomain,
+    dmu_tx_t *tx)
+{
+	fuid_domain_t searchnode, *findnode;
+	avl_index_t loc;
+
+	/*
+	 * If the dummy "nobody" domain then return an index of 0
+	 * to cause the created FUID to be a standard POSIX id
+	 * for the user nobody.
+	 */
+	if (domain[0] == '\0') {
+		*retdomain = "";
+		return (0);
+	}
+
+	searchnode.f_ksid = ksid_lookupdomain(domain);
+	if (retdomain) {
+		*retdomain = searchnode.f_ksid->kd_name;
+	}
+	if (!zfsvfs->z_fuid_loaded)
+		zfs_fuid_init(zfsvfs, tx);
+
+	rw_enter(&zfsvfs->z_fuid_lock, RW_READER);
+	findnode = avl_find(&zfsvfs->z_fuid_domain, &searchnode, &loc);
+	rw_exit(&zfsvfs->z_fuid_lock);
+
+	if (findnode) {
+		ksiddomain_rele(searchnode.f_ksid);
+		return (findnode->f_idx);
+	} else {
+		fuid_domain_t *domnode;
+		nvlist_t *nvp;
+		nvlist_t **fuids;
+		uint64_t retidx;
+		size_t nvsize = 0;
+		char *packed;
+		dmu_buf_t *db;
+		int i = 0;
+
+		domnode = kmem_alloc(sizeof (fuid_domain_t), KM_SLEEP);
+		domnode->f_ksid = searchnode.f_ksid;
+
+		rw_enter(&zfsvfs->z_fuid_lock, RW_WRITER);
+		retidx = domnode->f_idx = avl_numnodes(&zfsvfs->z_fuid_idx) + 1;
+
+		avl_add(&zfsvfs->z_fuid_domain, domnode);
+		avl_add(&zfsvfs->z_fuid_idx, domnode);
+		/*
+		 * Now resync the on-disk nvlist.
+		 */
+		VERIFY(nvlist_alloc(&nvp, NV_UNIQUE_NAME, KM_SLEEP) == 0);
+
+		domnode = avl_first(&zfsvfs->z_fuid_domain);
+		fuids = kmem_alloc(retidx * sizeof (void *), KM_SLEEP);
+		while (domnode) {
+			VERIFY(nvlist_alloc(&fuids[i],
+			    NV_UNIQUE_NAME, KM_SLEEP) == 0);
+			VERIFY(nvlist_add_uint64(fuids[i], FUID_IDX,
+			    domnode->f_idx) == 0);
+			VERIFY(nvlist_add_uint64(fuids[i],
+			    FUID_OFFSET, 0) == 0);
+			VERIFY(nvlist_add_string(fuids[i++], FUID_DOMAIN,
+			    domnode->f_ksid->kd_name) == 0);
+			domnode = AVL_NEXT(&zfsvfs->z_fuid_domain, domnode);
+		}
+		VERIFY(nvlist_add_nvlist_array(nvp, FUID_NVP_ARRAY,
+		    fuids, retidx) == 0);
+		for (i = 0; i != retidx; i++)
+			nvlist_free(fuids[i]);
+		kmem_free(fuids, retidx * sizeof (void *));
+		VERIFY(nvlist_size(nvp, &nvsize, NV_ENCODE_XDR) == 0);
+		packed = kmem_alloc(nvsize, KM_SLEEP);
+		VERIFY(nvlist_pack(nvp, &packed, &nvsize,
+		    NV_ENCODE_XDR, KM_SLEEP) == 0);
+		nvlist_free(nvp);
+		zfsvfs->z_fuid_size = nvsize;
+		dmu_write(zfsvfs->z_os, zfsvfs->z_fuid_obj, 0,
+		    zfsvfs->z_fuid_size, packed, tx);
+		kmem_free(packed, zfsvfs->z_fuid_size);
+		VERIFY(0 == dmu_bonus_hold(zfsvfs->z_os, zfsvfs->z_fuid_obj,
+		    FTAG, &db));
+		dmu_buf_will_dirty(db, tx);
+		*(uint64_t *)db->db_data = zfsvfs->z_fuid_size;
+		dmu_buf_rele(db, FTAG);
+
+		rw_exit(&zfsvfs->z_fuid_lock);
+		return (retidx);
+	}
+}
+
+/*
+ * Query domain table by index, returning domain string
+ *
+ * Returns a pointer from an avl node of the domain string.
+ *
+ */
+static char *
+zfs_fuid_find_by_idx(zfsvfs_t *zfsvfs, uint32_t idx)
+{
+	char *domain;
+
+	if (idx == 0 || !zfsvfs->z_use_fuids)
+		return (NULL);
+
+	if (!zfsvfs->z_fuid_loaded)
+		zfs_fuid_init(zfsvfs, NULL);
+
+	rw_enter(&zfsvfs->z_fuid_lock, RW_READER);
+	domain = zfs_fuid_idx_domain(&zfsvfs->z_fuid_idx, idx);
+	rw_exit(&zfsvfs->z_fuid_lock);
+
+	ASSERT(domain);
+	return (domain);
+}
+
+void
+zfs_fuid_map_ids(znode_t *zp, cred_t *cr, uid_t *uidp, uid_t *gidp)
+{
+	*uidp = zfs_fuid_map_id(zp->z_zfsvfs, zp->z_phys->zp_uid,
+	    cr, ZFS_OWNER);
+	*gidp = zfs_fuid_map_id(zp->z_zfsvfs, zp->z_phys->zp_gid,
+	    cr, ZFS_GROUP);
+}
+
+uid_t
+zfs_fuid_map_id(zfsvfs_t *zfsvfs, uint64_t fuid,
+    cred_t *cr, zfs_fuid_type_t type)
+{
+	uint32_t index = FUID_INDEX(fuid);
+	char *domain;
+	uid_t id;
+
+	if (index == 0)
+		return (fuid);
+
+	domain = zfs_fuid_find_by_idx(zfsvfs, index);
+	ASSERT(domain != NULL);
+
+	if (type == ZFS_OWNER || type == ZFS_ACE_USER) {
+		(void) kidmap_getuidbysid(crgetzone(cr), domain,
+		    FUID_RID(fuid), &id);
+	} else {
+		(void) kidmap_getgidbysid(crgetzone(cr), domain,
+		    FUID_RID(fuid), &id);
+	}
+	return (id);
+}
+
+/*
+ * Add a FUID node to the list of fuid's being created for this
+ * ACL
+ *
+ * If ACL has multiple domains, then keep only one copy of each unique
+ * domain.
+ */
+static void
+zfs_fuid_node_add(zfs_fuid_info_t **fuidpp, const char *domain, uint32_t rid,
+    uint64_t idx, uint64_t id, zfs_fuid_type_t type)
+{
+	zfs_fuid_t *fuid;
+	zfs_fuid_domain_t *fuid_domain;
+	zfs_fuid_info_t *fuidp;
+	uint64_t fuididx;
+	boolean_t found = B_FALSE;
+
+	if (*fuidpp == NULL)
+		*fuidpp = zfs_fuid_info_alloc();
+
+	fuidp = *fuidpp;
+	/*
+	 * First find fuid domain index in linked list
+	 *
+	 * If one isn't found then create an entry.
+	 */
+
+	for (fuididx = 1, fuid_domain = list_head(&fuidp->z_domains);
+	    fuid_domain; fuid_domain = list_next(&fuidp->z_domains,
+	    fuid_domain), fuididx++) {
+		if (idx == fuid_domain->z_domidx) {
+			found = B_TRUE;
+			break;
+		}
+	}
+
+	if (!found) {
+		fuid_domain = kmem_alloc(sizeof (zfs_fuid_domain_t), KM_SLEEP);
+		fuid_domain->z_domain = domain;
+		fuid_domain->z_domidx = idx;
+		list_insert_tail(&fuidp->z_domains, fuid_domain);
+		fuidp->z_domain_str_sz += strlen(domain) + 1;
+		fuidp->z_domain_cnt++;
+	}
+
+	if (type == ZFS_ACE_USER || type == ZFS_ACE_GROUP) {
+		/*
+		 * Now allocate fuid entry and add it on the end of the list
+		 */
+
+		fuid = kmem_alloc(sizeof (zfs_fuid_t), KM_SLEEP);
+		fuid->z_id = id;
+		fuid->z_domidx = idx;
+		fuid->z_logfuid = FUID_ENCODE(fuididx, rid);
+
+		list_insert_tail(&fuidp->z_fuids, fuid);
+		fuidp->z_fuid_cnt++;
+	} else {
+		if (type == ZFS_OWNER)
+			fuidp->z_fuid_owner = FUID_ENCODE(fuididx, rid);
+		else
+			fuidp->z_fuid_group = FUID_ENCODE(fuididx, rid);
+	}
+}
+
+/*
+ * Create a file system FUID, based on information in the users cred
+ */
+uint64_t
+zfs_fuid_create_cred(zfsvfs_t *zfsvfs, zfs_fuid_type_t type,
+    dmu_tx_t *tx, cred_t *cr, zfs_fuid_info_t **fuidp)
+{
+	uint64_t	idx;
+	ksid_t		*ksid;
+	uint32_t	rid;
+	char 		*kdomain;
+	const char	*domain;
+	uid_t		id;
+
+	VERIFY(type == ZFS_OWNER || type == ZFS_GROUP);
+
+	if (type == ZFS_OWNER)
+		id = crgetuid(cr);
+	else
+		id = crgetgid(cr);
+
+	if (!zfsvfs->z_use_fuids || !IS_EPHEMERAL(id))
+		return ((uint64_t)id);
+
+	ksid = crgetsid(cr, (type == ZFS_OWNER) ? KSID_OWNER : KSID_GROUP);
+
+	VERIFY(ksid != NULL);
+	rid = ksid_getrid(ksid);
+	domain = ksid_getdomain(ksid);
+
+	idx = zfs_fuid_find_by_domain(zfsvfs, domain, &kdomain, tx);
+
+	zfs_fuid_node_add(fuidp, kdomain, rid, idx, id, type);
+
+	return (FUID_ENCODE(idx, rid));
+}
+
+/*
+ * Create a file system FUID for an ACL ace
+ * or a chown/chgrp of the file.
+ * This is similar to zfs_fuid_create_cred, except that
+ * we can't find the domain + rid information in the
+ * cred.  Instead we have to query Winchester for the
+ * domain and rid.
+ *
+ * During replay operations the domain+rid information is
+ * found in the zfs_fuid_info_t that the replay code has
+ * attached to the zfsvfs of the file system.
+ */
+uint64_t
+zfs_fuid_create(zfsvfs_t *zfsvfs, uint64_t id, cred_t *cr,
+    zfs_fuid_type_t type, dmu_tx_t *tx, zfs_fuid_info_t **fuidpp)
+{
+	const char *domain;
+	char *kdomain;
+	uint32_t fuid_idx = FUID_INDEX(id);
+	uint32_t rid;
+	idmap_stat status;
+	uint64_t idx;
+	boolean_t is_replay = (zfsvfs->z_assign >= TXG_INITIAL);
+	zfs_fuid_t *zfuid = NULL;
+	zfs_fuid_info_t *fuidp;
+
+	/*
+	 * If POSIX ID, or entry is already a FUID then
+	 * just return the id
+	 *
+	 * We may also be handed an already FUID'ized id via
+	 * chmod.
+	 */
+
+	if (!zfsvfs->z_use_fuids || !IS_EPHEMERAL(id) || fuid_idx != 0)
+		return (id);
+
+	if (is_replay) {
+		fuidp = zfsvfs->z_fuid_replay;
+
+		/*
+		 * If we are passed an ephemeral id, but no
+		 * fuid_info was logged then return NOBODY.
+		 * This is most likely a result of idmap service
+		 * not being available.
+		 */
+		if (fuidp == NULL)
+			return (UID_NOBODY);
+
+		switch (type) {
+		case ZFS_ACE_USER:
+		case ZFS_ACE_GROUP:
+			zfuid = list_head(&fuidp->z_fuids);
+			rid = FUID_RID(zfuid->z_logfuid);
+			idx = FUID_INDEX(zfuid->z_logfuid);
+			break;
+		case ZFS_OWNER:
+			rid = FUID_RID(fuidp->z_fuid_owner);
+			idx = FUID_INDEX(fuidp->z_fuid_owner);
+			break;
+		case ZFS_GROUP:
+			rid = FUID_RID(fuidp->z_fuid_group);
+			idx = FUID_INDEX(fuidp->z_fuid_group);
+			break;
+		};
+		domain = fuidp->z_domain_table[idx -1];
+	} else {
+		if (type == ZFS_OWNER || type == ZFS_ACE_USER)
+			status = kidmap_getsidbyuid(crgetzone(cr), id,
+			    &domain, &rid);
+		else
+			status = kidmap_getsidbygid(crgetzone(cr), id,
+			    &domain, &rid);
+
+		if (status != 0) {
+			/*
+			 * When returning nobody we will need to
+			 * make a dummy fuid table entry for logging
+			 * purposes.
+			 */
+			rid = UID_NOBODY;
+			domain = "";
+		}
+	}
+
+	idx = zfs_fuid_find_by_domain(zfsvfs, domain, &kdomain, tx);
+
+	if (!is_replay)
+		zfs_fuid_node_add(fuidpp, kdomain, rid, idx, id, type);
+	else if (zfuid != NULL) {
+		list_remove(&fuidp->z_fuids, zfuid);
+		kmem_free(zfuid, sizeof (zfs_fuid_t));
+	}
+	return (FUID_ENCODE(idx, rid));
+}
+
+void
+zfs_fuid_destroy(zfsvfs_t *zfsvfs)
+{
+	rw_enter(&zfsvfs->z_fuid_lock, RW_WRITER);
+	if (!zfsvfs->z_fuid_loaded) {
+		rw_exit(&zfsvfs->z_fuid_lock);
+		return;
+	}
+	zfs_fuid_table_destroy(&zfsvfs->z_fuid_idx, &zfsvfs->z_fuid_domain);
+	rw_exit(&zfsvfs->z_fuid_lock);
+}
+
+/*
+ * Allocate zfs_fuid_info for tracking FUIDs created during
+ * zfs_mknode, VOP_SETATTR() or VOP_SETSECATTR()
+ */
+zfs_fuid_info_t *
+zfs_fuid_info_alloc(void)
+{
+	zfs_fuid_info_t *fuidp;
+
+	fuidp = kmem_zalloc(sizeof (zfs_fuid_info_t), KM_SLEEP);
+	list_create(&fuidp->z_domains, sizeof (zfs_fuid_domain_t),
+	    offsetof(zfs_fuid_domain_t, z_next));
+	list_create(&fuidp->z_fuids, sizeof (zfs_fuid_t),
+	    offsetof(zfs_fuid_t, z_next));
+	return (fuidp);
+}
+
+/*
+ * Release all memory associated with zfs_fuid_info_t
+ */
+void
+zfs_fuid_info_free(zfs_fuid_info_t *fuidp)
+{
+	zfs_fuid_t *zfuid;
+	zfs_fuid_domain_t *zdomain;
+
+	while ((zfuid = list_head(&fuidp->z_fuids)) != NULL) {
+		list_remove(&fuidp->z_fuids, zfuid);
+		kmem_free(zfuid, sizeof (zfs_fuid_t));
+	}
+
+	if (fuidp->z_domain_table != NULL)
+		kmem_free(fuidp->z_domain_table,
+		    (sizeof (char **)) * fuidp->z_domain_cnt);
+
+	while ((zdomain = list_head(&fuidp->z_domains)) != NULL) {
+		list_remove(&fuidp->z_domains, zdomain);
+		kmem_free(zdomain, sizeof (zfs_fuid_domain_t));
+	}
+
+	kmem_free(fuidp, sizeof (zfs_fuid_info_t));
+}
+
+/*
+ * Check to see if id is a groupmember.  If cred
+ * has ksid info then sidlist is checked first
+ * and if still not found then POSIX groups are checked
+ *
+ * Will use a straight FUID compare when possible.
+ */
+boolean_t
+zfs_groupmember(zfsvfs_t *zfsvfs, uint64_t id, cred_t *cr)
+{
+	ksid_t		*ksid = crgetsid(cr, KSID_GROUP);
+	uid_t		gid;
+
+	if (ksid) {
+		int 		i;
+		ksid_t		*ksid_groups;
+		ksidlist_t	*ksidlist = crgetsidlist(cr);
+		uint32_t	idx = FUID_INDEX(id);
+		uint32_t	rid = FUID_RID(id);
+
+		ASSERT(ksidlist);
+		ksid_groups = ksidlist->ksl_sids;
+
+		for (i = 0; i != ksidlist->ksl_nsid; i++) {
+			if (idx == 0) {
+				if (id != IDMAP_WK_CREATOR_GROUP_GID &&
+				    id == ksid_groups[i].ks_id) {
+					return (B_TRUE);
+				}
+			} else {
+				char *domain;
+
+				domain = zfs_fuid_find_by_idx(zfsvfs, idx);
+				ASSERT(domain != NULL);
+
+				if (strcmp(domain,
+				    IDMAP_WK_CREATOR_SID_AUTHORITY) == 0)
+					return (B_FALSE);
+
+				if ((strcmp(domain,
+				    ksid_groups[i].ks_domain->kd_name) == 0) &&
+				    rid == ksid_groups[i].ks_rid)
+					return (B_TRUE);
+			}
+		}
+	}
+
+	/*
+	 * Not found in ksidlist, check posix groups
+	 */
+	gid = zfs_fuid_map_id(zfsvfs, id, cr, ZFS_GROUP);
+	return (groupmember(gid, cr));
+}
+#endif
--- a/zfs/lib/libdmu-ctl/zfs_ioctl.c
+++ b/zfs/lib/libdmu-ctl/zfs_ioctl.c
--- a/zfs/lib/libdmu-ctl/zfs_log.c
+++ b/zfs/lib/libdmu-ctl/zfs_log.c
@ -0,0 +1,693 @@
+/*
+ * CDDL HEADER START
+ *
+ * The contents of this file are subject to the terms of the
+ * Common Development and Distribution License (the "License").
+ * You may not use this file except in compliance with the License.
+ *
+ * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
+ * or http://www.opensolaris.org/os/licensing.
+ * See the License for the specific language governing permissions
+ * and limitations under the License.
+ *
+ * When distributing Covered Code, include this CDDL HEADER in each
+ * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
+ * If applicable, add the following below this CDDL HEADER, with the
+ * fields enclosed by brackets "[]" replaced with your own identifying
+ * information: Portions Copyright [yyyy] [name of copyright owner]
+ *
+ * CDDL HEADER END
+ */
+/*
+ * Copyright 2008 Sun Microsystems, Inc.  All rights reserved.
+ * Use is subject to license terms.
+ */
+
+#pragma ident	"@(#)zfs_log.c	1.13	08/04/09 SMI"
+
+#include <sys/types.h>
+#include <sys/param.h>
+#include <sys/systm.h>
+#include <sys/sysmacros.h>
+#include <sys/cmn_err.h>
+#include <sys/kmem.h>
+#include <sys/thread.h>
+#include <sys/file.h>
+#include <sys/vfs.h>
+#include <sys/zfs_znode.h>
+#include <sys/zfs_dir.h>
+#include <sys/zil.h>
+#include <sys/zil_impl.h>
+#include <sys/byteorder.h>
+#include <sys/policy.h>
+#include <sys/stat.h>
+#include <sys/mode.h>
+#include <sys/acl.h>
+#include <sys/dmu.h>
+#include <sys/spa.h>
+#include <sys/zfs_fuid.h>
+#include <sys/ddi.h>
+
+/*
+ * All the functions in this file are used to construct the log entries
+ * to record transactions. They allocate * an intent log transaction
+ * structure (itx_t) and save within it all the information necessary to
+ * possibly replay the transaction. The itx is then assigned a sequence
+ * number and inserted in the in-memory list anchored in the zilog.
+ */
+
+int
+zfs_log_create_txtype(zil_create_t type, vsecattr_t *vsecp, vattr_t *vap)
+{
+	int isxvattr = (vap->va_mask & AT_XVATTR);
+	switch (type) {
+	case Z_FILE:
+		if (vsecp == NULL && !isxvattr)
+			return (TX_CREATE);
+		if (vsecp && isxvattr)
+			return (TX_CREATE_ACL_ATTR);
+		if (vsecp)
+			return (TX_CREATE_ACL);
+		else
+			return (TX_CREATE_ATTR);
+		/*NOTREACHED*/
+	case Z_DIR:
+		if (vsecp == NULL && !isxvattr)
+			return (TX_MKDIR);
+		if (vsecp && isxvattr)
+			return (TX_MKDIR_ACL_ATTR);
+		if (vsecp)
+			return (TX_MKDIR_ACL);
+		else
+			return (TX_MKDIR_ATTR);
+	case Z_XATTRDIR:
+		return (TX_MKXATTR);
+	}
+	ASSERT(0);
+	return (TX_MAX_TYPE);
+}
+
+/*
+ * build up the log data necessary for logging xvattr_t
+ * First lr_attr_t is initialized.  following the lr_attr_t
+ * is the mapsize and attribute bitmap copied from the xvattr_t.
+ * Following the bitmap and bitmapsize two 64 bit words are reserved
+ * for the create time which may be set.  Following the create time
+ * records a single 64 bit integer which has the bits to set on
+ * replay for the xvattr.
+ */
+static void
+zfs_log_xvattr(lr_attr_t *lrattr, xvattr_t *xvap)
+{
+	uint32_t	*bitmap;
+	uint64_t	*attrs;
+	uint64_t	*crtime;
+	xoptattr_t	*xoap;
+	void		*scanstamp;
+	int		i;
+
+	xoap = xva_getxoptattr(xvap);
+	ASSERT(xoap);
+
+	lrattr->lr_attr_masksize = xvap->xva_mapsize;
+	bitmap = &lrattr->lr_attr_bitmap;
+	for (i = 0; i != xvap->xva_mapsize; i++, bitmap++) {
+		*bitmap = xvap->xva_reqattrmap[i];
+	}
+
+	/* Now pack the attributes up in a single uint64_t */
+	attrs = (uint64_t *)bitmap;
+	crtime = attrs + 1;
+	scanstamp = (caddr_t)(crtime + 2);
+	*attrs = 0;
+	if (XVA_ISSET_REQ(xvap, XAT_READONLY))
+		*attrs |= (xoap->xoa_readonly == 0) ? 0 :
+		    XAT0_READONLY;
+	if (XVA_ISSET_REQ(xvap, XAT_HIDDEN))
+		*attrs |= (xoap->xoa_hidden == 0) ? 0 :
+		    XAT0_HIDDEN;
+	if (XVA_ISSET_REQ(xvap, XAT_SYSTEM))
+		*attrs |= (xoap->xoa_system == 0) ? 0 :
+		    XAT0_SYSTEM;
+	if (XVA_ISSET_REQ(xvap, XAT_ARCHIVE))
+		*attrs |= (xoap->xoa_archive == 0) ? 0 :
+		    XAT0_ARCHIVE;
+	if (XVA_ISSET_REQ(xvap, XAT_IMMUTABLE))
+		*attrs |= (xoap->xoa_immutable == 0) ? 0 :
+		    XAT0_IMMUTABLE;
+	if (XVA_ISSET_REQ(xvap, XAT_NOUNLINK))
+		*attrs |= (xoap->xoa_nounlink == 0) ? 0 :
+		    XAT0_NOUNLINK;
+	if (XVA_ISSET_REQ(xvap, XAT_APPENDONLY))
+		*attrs |= (xoap->xoa_appendonly == 0) ? 0 :
+		    XAT0_APPENDONLY;
+	if (XVA_ISSET_REQ(xvap, XAT_OPAQUE))
+		*attrs |= (xoap->xoa_opaque == 0) ? 0 :
+		    XAT0_APPENDONLY;
+	if (XVA_ISSET_REQ(xvap, XAT_NODUMP))
+		*attrs |= (xoap->xoa_nodump == 0) ? 0 :
+		    XAT0_NODUMP;
+	if (XVA_ISSET_REQ(xvap, XAT_AV_QUARANTINED))
+		*attrs |= (xoap->xoa_av_quarantined == 0) ? 0 :
+		    XAT0_AV_QUARANTINED;
+	if (XVA_ISSET_REQ(xvap, XAT_AV_MODIFIED))
+		*attrs |= (xoap->xoa_av_modified == 0) ? 0 :
+		    XAT0_AV_MODIFIED;
+	if (XVA_ISSET_REQ(xvap, XAT_CREATETIME))
+		ZFS_TIME_ENCODE(&xoap->xoa_createtime, crtime);
+	if (XVA_ISSET_REQ(xvap, XAT_AV_SCANSTAMP))
+		bcopy(xoap->xoa_av_scanstamp, scanstamp, AV_SCANSTAMP_SZ);
+}
+
+static void *
+zfs_log_fuid_ids(zfs_fuid_info_t *fuidp, void *start)
+{
+	zfs_fuid_t *zfuid;
+	uint64_t *fuidloc = start;
+
+	/* First copy in the ACE FUIDs */
+	for (zfuid = list_head(&fuidp->z_fuids); zfuid;
+	    zfuid = list_next(&fuidp->z_fuids, zfuid)) {
+		*fuidloc++ = zfuid->z_logfuid;
+	}
+	return (fuidloc);
+}
+
+
+static void *
+zfs_log_fuid_domains(zfs_fuid_info_t *fuidp, void *start)
+{
+	zfs_fuid_domain_t *zdomain;
+
+	/* now copy in the domain info, if any */
+	if (fuidp->z_domain_str_sz != 0) {
+		for (zdomain = list_head(&fuidp->z_domains); zdomain;
+		    zdomain = list_next(&fuidp->z_domains, zdomain)) {
+			bcopy((void *)zdomain->z_domain, start,
+			    strlen(zdomain->z_domain) + 1);
+			start = (caddr_t)start +
+			    strlen(zdomain->z_domain) + 1;
+		}
+	}
+	return (start);
+}
+
+/*
+ * zfs_log_create() is used to handle TX_CREATE, TX_CREATE_ATTR, TX_MKDIR,
+ * TX_MKDIR_ATTR and TX_MKXATTR
+ * transactions.
+ *
+ * TX_CREATE and TX_MKDIR are standard creates, but they may have FUID
+ * domain information appended prior to the name.  In this case the
+ * uid/gid in the log record will be a log centric FUID.
+ *
+ * TX_CREATE_ACL_ATTR and TX_MKDIR_ACL_ATTR handle special creates that
+ * may contain attributes, ACL and optional fuid information.
+ *
+ * TX_CREATE_ACL and TX_MKDIR_ACL handle special creates that specify
+ * and ACL and normal users/groups in the ACEs.
+ *
+ * There may be an optional xvattr attribute information similar
+ * to zfs_log_setattr.
+ *
+ * Also, after the file name "domain" strings may be appended.
+ */
+void
+zfs_log_create(zilog_t *zilog, dmu_tx_t *tx, uint64_t txtype,
+    znode_t *dzp, znode_t *zp, char *name, vsecattr_t *vsecp,
+    zfs_fuid_info_t *fuidp, vattr_t *vap)
+{
+	itx_t *itx;
+	uint64_t seq;
+	lr_create_t *lr;
+	lr_acl_create_t *lracl;
+	size_t aclsize;
+	size_t xvatsize = 0;
+	size_t txsize;
+	xvattr_t *xvap = (xvattr_t *)vap;
+	void *end;
+	size_t lrsize;
+
+	size_t namesize = strlen(name) + 1;
+	size_t fuidsz = 0;
+
+	if (zilog == NULL)
+		return;
+
+	/*
+	 * If we have FUIDs present then add in space for
+	 * domains and ACE fuid's if any.
+	 */
+	if (fuidp) {
+		fuidsz += fuidp->z_domain_str_sz;
+		fuidsz += fuidp->z_fuid_cnt * sizeof (uint64_t);
+	}
+
+	if (vap->va_mask & AT_XVATTR)
+		xvatsize = ZIL_XVAT_SIZE(xvap->xva_mapsize);
+
+	if ((int)txtype == TX_CREATE_ATTR || (int)txtype == TX_MKDIR_ATTR ||
+	    (int)txtype == TX_CREATE || (int)txtype == TX_MKDIR ||
+	    (int)txtype == TX_MKXATTR) {
+		txsize = sizeof (*lr) + namesize + fuidsz + xvatsize;
+		lrsize = sizeof (*lr);
+	} else {
+		aclsize = (vsecp) ? vsecp->vsa_aclentsz : 0;
+		txsize =
+		    sizeof (lr_acl_create_t) + namesize + fuidsz +
+		    ZIL_ACE_LENGTH(aclsize) + xvatsize;
+		lrsize = sizeof (lr_acl_create_t);
+	}
+
+	itx = zil_itx_create(txtype, txsize);
+
+	lr = (lr_create_t *)&itx->itx_lr;
+	lr->lr_doid = dzp->z_id;
+	lr->lr_foid = zp->z_id;
+	lr->lr_mode = zp->z_phys->zp_mode;
+	if (!IS_EPHEMERAL(zp->z_phys->zp_uid)) {
+		lr->lr_uid = (uint64_t)zp->z_phys->zp_uid;
+	} else {
+		lr->lr_uid = fuidp->z_fuid_owner;
+	}
+	if (!IS_EPHEMERAL(zp->z_phys->zp_gid)) {
+		lr->lr_gid = (uint64_t)zp->z_phys->zp_gid;
+	} else {
+		lr->lr_gid = fuidp->z_fuid_group;
+	}
+	lr->lr_gen = zp->z_phys->zp_gen;
+	lr->lr_crtime[0] = zp->z_phys->zp_crtime[0];
+	lr->lr_crtime[1] = zp->z_phys->zp_crtime[1];
+	lr->lr_rdev = zp->z_phys->zp_rdev;
+
+	/*
+	 * Fill in xvattr info if any
+	 */
+	if (vap->va_mask & AT_XVATTR) {
+		zfs_log_xvattr((lr_attr_t *)((caddr_t)lr + lrsize), xvap);
+		end = (caddr_t)lr + lrsize + xvatsize;
+	} else {
+		end = (caddr_t)lr + lrsize;
+	}
+
+	/* Now fill in any ACL info */
+
+	if (vsecp) {
+		lracl = (lr_acl_create_t *)&itx->itx_lr;
+		lracl->lr_aclcnt = vsecp->vsa_aclcnt;
+		lracl->lr_acl_bytes = aclsize;
+		lracl->lr_domcnt = fuidp ? fuidp->z_domain_cnt : 0;
+		lracl->lr_fuidcnt  = fuidp ? fuidp->z_fuid_cnt : 0;
+		if (vsecp->vsa_aclflags & VSA_ACE_ACLFLAGS)
+			lracl->lr_acl_flags = (uint64_t)vsecp->vsa_aclflags;
+		else
+			lracl->lr_acl_flags = 0;
+
+		bcopy(vsecp->vsa_aclentp, end, aclsize);
+		end = (caddr_t)end + ZIL_ACE_LENGTH(aclsize);
+	}
+
+	/* drop in FUID info */
+	if (fuidp) {
+		end = zfs_log_fuid_ids(fuidp, end);
+		end = zfs_log_fuid_domains(fuidp, end);
+	}
+	/*
+	 * Now place file name in log record
+	 */
+	bcopy(name, end, namesize);
+
+	seq = zil_itx_assign(zilog, itx, tx);
+	dzp->z_last_itx = seq;
+	zp->z_last_itx = seq;
+}
+
+/*
+ * zfs_log_remove() handles both TX_REMOVE and TX_RMDIR transactions.
+ */
+void
+zfs_log_remove(zilog_t *zilog, dmu_tx_t *tx, uint64_t txtype,
+	znode_t *dzp, char *name)
+{
+	itx_t *itx;
+	uint64_t seq;
+	lr_remove_t *lr;
+	size_t namesize = strlen(name) + 1;
+
+	if (zilog == NULL)
+		return;
+
+	itx = zil_itx_create(txtype, sizeof (*lr) + namesize);
+	lr = (lr_remove_t *)&itx->itx_lr;
+	lr->lr_doid = dzp->z_id;
+	bcopy(name, (char *)(lr + 1), namesize);
+
+	seq = zil_itx_assign(zilog, itx, tx);
+	dzp->z_last_itx = seq;
+}
+
+/*
+ * zfs_log_link() handles TX_LINK transactions.
+ */
+void
+zfs_log_link(zilog_t *zilog, dmu_tx_t *tx, uint64_t txtype,
+	znode_t *dzp, znode_t *zp, char *name)
+{
+	itx_t *itx;
+	uint64_t seq;
+	lr_link_t *lr;
+	size_t namesize = strlen(name) + 1;
+
+	if (zilog == NULL)
+		return;
+
+	itx = zil_itx_create(txtype, sizeof (*lr) + namesize);
+	lr = (lr_link_t *)&itx->itx_lr;
+	lr->lr_doid = dzp->z_id;
+	lr->lr_link_obj = zp->z_id;
+	bcopy(name, (char *)(lr + 1), namesize);
+
+	seq = zil_itx_assign(zilog, itx, tx);
+	dzp->z_last_itx = seq;
+	zp->z_last_itx = seq;
+}
+
+/*
+ * zfs_log_symlink() handles TX_SYMLINK transactions.
+ */
+void
+zfs_log_symlink(zilog_t *zilog, dmu_tx_t *tx, uint64_t txtype,
+    znode_t *dzp, znode_t *zp, char *name, char *link)
+{
+	itx_t *itx;
+	uint64_t seq;
+	lr_create_t *lr;
+	size_t namesize = strlen(name) + 1;
+	size_t linksize = strlen(link) + 1;
+
+	if (zilog == NULL)
+		return;
+
+	itx = zil_itx_create(txtype, sizeof (*lr) + namesize + linksize);
+	lr = (lr_create_t *)&itx->itx_lr;
+	lr->lr_doid = dzp->z_id;
+	lr->lr_foid = zp->z_id;
+	lr->lr_mode = zp->z_phys->zp_mode;
+	lr->lr_uid = zp->z_phys->zp_uid;
+	lr->lr_gid = zp->z_phys->zp_gid;
+	lr->lr_gen = zp->z_phys->zp_gen;
+	lr->lr_crtime[0] = zp->z_phys->zp_crtime[0];
+	lr->lr_crtime[1] = zp->z_phys->zp_crtime[1];
+	bcopy(name, (char *)(lr + 1), namesize);
+	bcopy(link, (char *)(lr + 1) + namesize, linksize);
+
+	seq = zil_itx_assign(zilog, itx, tx);
+	dzp->z_last_itx = seq;
+	zp->z_last_itx = seq;
+}
+
+/*
+ * zfs_log_rename() handles TX_RENAME transactions.
+ */
+void
+zfs_log_rename(zilog_t *zilog, dmu_tx_t *tx, uint64_t txtype,
+	znode_t *sdzp, char *sname, znode_t *tdzp, char *dname, znode_t *szp)
+{
+	itx_t *itx;
+	uint64_t seq;
+	lr_rename_t *lr;
+	size_t snamesize = strlen(sname) + 1;
+	size_t dnamesize = strlen(dname) + 1;
+
+	if (zilog == NULL)
+		return;
+
+	itx = zil_itx_create(txtype, sizeof (*lr) + snamesize + dnamesize);
+	lr = (lr_rename_t *)&itx->itx_lr;
+	lr->lr_sdoid = sdzp->z_id;
+	lr->lr_tdoid = tdzp->z_id;
+	bcopy(sname, (char *)(lr + 1), snamesize);
+	bcopy(dname, (char *)(lr + 1) + snamesize, dnamesize);
+
+	seq = zil_itx_assign(zilog, itx, tx);
+	sdzp->z_last_itx = seq;
+	tdzp->z_last_itx = seq;
+	szp->z_last_itx = seq;
+}
+
+/*
+ * zfs_log_write() handles TX_WRITE transactions.
+ */
+ssize_t zfs_immediate_write_sz = 32768;
+
+#define	ZIL_MAX_LOG_DATA (SPA_MAXBLOCKSIZE - sizeof (zil_trailer_t) - \
+    sizeof (lr_write_t))
+
+void
+zfs_log_write(zilog_t *zilog, dmu_tx_t *tx, int txtype,
+	znode_t *zp, offset_t off, ssize_t resid, int ioflag)
+{
+	itx_wr_state_t write_state;
+	boolean_t slogging;
+	uintptr_t fsync_cnt;
+
+	if (zilog == NULL || zp->z_unlinked)
+		return;
+
+	/*
+	 * Writes are handled in three different ways:
+	 *
+	 * WR_INDIRECT:
+	 *    If the write is greater than zfs_immediate_write_sz and there are
+	 *    no separate logs in this pool then later *if* we need to log the
+	 *    write then dmu_sync() is used to immediately write the block and
+	 *    its block pointer is put in the log record.
+	 * WR_COPIED:
+	 *    If we know we'll immediately be committing the
+	 *    transaction (FSYNC or FDSYNC), the we allocate a larger
+	 *    log record here for the data and copy the data in.
+	 * WR_NEED_COPY:
+	 *    Otherwise we don't allocate a buffer, and *if* we need to
+	 *    flush the write later then a buffer is allocated and
+	 *    we retrieve the data using the dmu.
+	 */
+	slogging = spa_has_slogs(zilog->zl_spa);
+	if (resid > zfs_immediate_write_sz && !slogging)
+		write_state = WR_INDIRECT;
+	else if (ioflag & (FSYNC | FDSYNC))
+		write_state = WR_COPIED;
+	else
+		write_state = WR_NEED_COPY;
+
+	if ((fsync_cnt = (uintptr_t)tsd_get(zfs_fsyncer_key)) != 0) {
+		(void) tsd_set(zfs_fsyncer_key, (void *)(fsync_cnt - 1));
+	}
+
+	while (resid) {
+		itx_t *itx;
+		lr_write_t *lr;
+		ssize_t len;
+
+		/*
+		 * If there are slogs and the write would overflow the largest
+		 * block, then because we don't want to use the main pool
+		 * to dmu_sync, we have to split the write.
+		 */
+		if (slogging && resid > ZIL_MAX_LOG_DATA)
+			len = SPA_MAXBLOCKSIZE >> 1;
+		else
+			len = resid;
+
+		itx = zil_itx_create(txtype, sizeof (*lr) +
+		    (write_state == WR_COPIED ? len : 0));
+		lr = (lr_write_t *)&itx->itx_lr;
+		if (write_state == WR_COPIED && dmu_read(zp->z_zfsvfs->z_os,
+		    zp->z_id, off, len, lr + 1) != 0) {
+			kmem_free(itx, offsetof(itx_t, itx_lr) +
+			    itx->itx_lr.lrc_reclen);
+			itx = zil_itx_create(txtype, sizeof (*lr));
+			lr = (lr_write_t *)&itx->itx_lr;
+			write_state = WR_NEED_COPY;
+		}
+
+		itx->itx_wr_state = write_state;
+		if (write_state == WR_NEED_COPY)
+			itx->itx_sod += len;
+		lr->lr_foid = zp->z_id;
+		lr->lr_offset = off;
+		lr->lr_length = len;
+		lr->lr_blkoff = 0;
+		BP_ZERO(&lr->lr_blkptr);
+
+		itx->itx_private = zp->z_zfsvfs;
+
+		if ((zp->z_sync_cnt != 0) || (fsync_cnt != 0) ||
+		    (ioflag & (FSYNC | FDSYNC)))
+			itx->itx_sync = B_TRUE;
+		else
+			itx->itx_sync = B_FALSE;
+
+		zp->z_last_itx = zil_itx_assign(zilog, itx, tx);
+
+		off += len;
+		resid -= len;
+	}
+}
+
+/*
+ * zfs_log_truncate() handles TX_TRUNCATE transactions.
+ */
+void
+zfs_log_truncate(zilog_t *zilog, dmu_tx_t *tx, int txtype,
+	znode_t *zp, uint64_t off, uint64_t len)
+{
+	itx_t *itx;
+	uint64_t seq;
+	lr_truncate_t *lr;
+
+	if (zilog == NULL || zp->z_unlinked)
+		return;
+
+	itx = zil_itx_create(txtype, sizeof (*lr));
+	lr = (lr_truncate_t *)&itx->itx_lr;
+	lr->lr_foid = zp->z_id;
+	lr->lr_offset = off;
+	lr->lr_length = len;
+
+	itx->itx_sync = (zp->z_sync_cnt != 0);
+	seq = zil_itx_assign(zilog, itx, tx);
+	zp->z_last_itx = seq;
+}
+
+/*
+ * zfs_log_setattr() handles TX_SETATTR transactions.
+ */
+void
+zfs_log_setattr(zilog_t *zilog, dmu_tx_t *tx, int txtype,
+	znode_t *zp, vattr_t *vap, uint_t mask_applied, zfs_fuid_info_t *fuidp)
+{
+	itx_t		*itx;
+	uint64_t	seq;
+	lr_setattr_t	*lr;
+	xvattr_t	*xvap = (xvattr_t *)vap;
+	size_t		recsize = sizeof (lr_setattr_t);
+	void		*start;
+
+
+	if (zilog == NULL || zp->z_unlinked)
+		return;
+
+	/*
+	 * If XVATTR set, then log record size needs to allow
+	 * for lr_attr_t + xvattr mask, mapsize and create time
+	 * plus actual attribute values
+	 */
+	if (vap->va_mask & AT_XVATTR)
+		recsize = sizeof (*lr) + ZIL_XVAT_SIZE(xvap->xva_mapsize);
+
+	if (fuidp)
+		recsize += fuidp->z_domain_str_sz;
+
+	itx = zil_itx_create(txtype, recsize);
+	lr = (lr_setattr_t *)&itx->itx_lr;
+	lr->lr_foid = zp->z_id;
+	lr->lr_mask = (uint64_t)mask_applied;
+	lr->lr_mode = (uint64_t)vap->va_mode;
+	if ((mask_applied & AT_UID) && IS_EPHEMERAL(vap->va_uid))
+		lr->lr_uid = fuidp->z_fuid_owner;
+	else
+		lr->lr_uid = (uint64_t)vap->va_uid;
+
+	if ((mask_applied & AT_GID) && IS_EPHEMERAL(vap->va_gid))
+		lr->lr_gid = fuidp->z_fuid_group;
+	else
+		lr->lr_gid = (uint64_t)vap->va_gid;
+
+	lr->lr_size = (uint64_t)vap->va_size;
+	ZFS_TIME_ENCODE(&vap->va_atime, lr->lr_atime);
+	ZFS_TIME_ENCODE(&vap->va_mtime, lr->lr_mtime);
+	start = (lr_setattr_t *)(lr + 1);
+	if (vap->va_mask & AT_XVATTR) {
+		zfs_log_xvattr((lr_attr_t *)start, xvap);
+		start = (caddr_t)start + ZIL_XVAT_SIZE(xvap->xva_mapsize);
+	}
+
+	/*
+	 * Now stick on domain information if any on end
+	 */
+
+	if (fuidp)
+		(void) zfs_log_fuid_domains(fuidp, start);
+
+	itx->itx_sync = (zp->z_sync_cnt != 0);
+	seq = zil_itx_assign(zilog, itx, tx);
+	zp->z_last_itx = seq;
+}
+
+/*
+ * zfs_log_acl() handles TX_ACL transactions.
+ */
+void
+zfs_log_acl(zilog_t *zilog, dmu_tx_t *tx, znode_t *zp,
+    vsecattr_t *vsecp, zfs_fuid_info_t *fuidp)
+{
+	itx_t *itx;
+	uint64_t seq;
+	lr_acl_v0_t *lrv0;
+	lr_acl_t *lr;
+	int txtype;
+	int lrsize;
+	size_t txsize;
+	size_t aclbytes = vsecp->vsa_aclentsz;
+
+	txtype = (zp->z_zfsvfs->z_version == ZPL_VERSION_INITIAL) ?
+	    TX_ACL_V0 : TX_ACL;
+
+	if (txtype == TX_ACL)
+		lrsize = sizeof (*lr);
+	else
+		lrsize = sizeof (*lrv0);
+
+	if (zilog == NULL || zp->z_unlinked)
+		return;
+
+	txsize = lrsize +
+	    ((txtype == TX_ACL) ? ZIL_ACE_LENGTH(aclbytes) : aclbytes) +
+	    (fuidp ? fuidp->z_domain_str_sz : 0) +
+	    sizeof (uint64) * (fuidp ? fuidp->z_fuid_cnt : 0);
+
+	itx = zil_itx_create(txtype, txsize);
+
+	lr = (lr_acl_t *)&itx->itx_lr;
+	lr->lr_foid = zp->z_id;
+	if (txtype == TX_ACL) {
+		lr->lr_acl_bytes = aclbytes;
+		lr->lr_domcnt = fuidp ? fuidp->z_domain_cnt : 0;
+		lr->lr_fuidcnt = fuidp ? fuidp->z_fuid_cnt : 0;
+		if (vsecp->vsa_mask & VSA_ACE_ACLFLAGS)
+			lr->lr_acl_flags = (uint64_t)vsecp->vsa_aclflags;
+		else
+			lr->lr_acl_flags = 0;
+	}
+	lr->lr_aclcnt = (uint64_t)vsecp->vsa_aclcnt;
+
+	if (txtype == TX_ACL_V0) {
+		lrv0 = (lr_acl_v0_t *)lr;
+		bcopy(vsecp->vsa_aclentp, (ace_t *)(lrv0 + 1), aclbytes);
+	} else {
+		void *start = (ace_t *)(lr + 1);
+
+		bcopy(vsecp->vsa_aclentp, start, aclbytes);
+
+		start = (caddr_t)start + ZIL_ACE_LENGTH(aclbytes);
+
+		if (fuidp) {
+			start = zfs_log_fuid_ids(fuidp, start);
+			(void) zfs_log_fuid_domains(fuidp, start);
+		}
+	}
+
+	itx->itx_sync = (zp->z_sync_cnt != 0);
+	seq = zil_itx_assign(zilog, itx, tx);
+	zp->z_last_itx = seq;
+}
--- a/zfs/lib/libdmu-ctl/zfs_replay.c
+++ b/zfs/lib/libdmu-ctl/zfs_replay.c
@ -0,0 +1,876 @@
+/*
+ * CDDL HEADER START
+ *
+ * The contents of this file are subject to the terms of the
+ * Common Development and Distribution License (the "License").
+ * You may not use this file except in compliance with the License.
+ *
+ * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
+ * or http://www.opensolaris.org/os/licensing.
+ * See the License for the specific language governing permissions
+ * and limitations under the License.
+ *
+ * When distributing Covered Code, include this CDDL HEADER in each
+ * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
+ * If applicable, add the following below this CDDL HEADER, with the
+ * fields enclosed by brackets "[]" replaced with your own identifying
+ * information: Portions Copyright [yyyy] [name of copyright owner]
+ *
+ * CDDL HEADER END
+ */
+/*
+ * Copyright 2008 Sun Microsystems, Inc.  All rights reserved.
+ * Use is subject to license terms.
+ */
+
+#pragma ident	"@(#)zfs_replay.c	1.7	08/01/14 SMI"
+
+#include <sys/types.h>
+#include <sys/param.h>
+#include <sys/systm.h>
+#include <sys/sysmacros.h>
+#include <sys/cmn_err.h>
+#include <sys/kmem.h>
+#include <sys/thread.h>
+#include <sys/file.h>
+#include <sys/fcntl.h>
+#include <sys/vfs.h>
+#include <sys/fs/zfs.h>
+#include <sys/zfs_znode.h>
+#include <sys/zfs_dir.h>
+#include <sys/zfs_acl.h>
+#include <sys/zfs_fuid.h>
+#include <sys/spa.h>
+#include <sys/zil.h>
+#include <sys/byteorder.h>
+#include <sys/stat.h>
+#include <sys/mode.h>
+#include <sys/acl.h>
+#include <sys/atomic.h>
+#include <sys/cred.h>
+
+/*
+ * Functions to replay ZFS intent log (ZIL) records
+ * The functions are called through a function vector (zfs_replay_vector)
+ * which is indexed by the transaction type.
+ */
+
+static void
+zfs_init_vattr(vattr_t *vap, uint64_t mask, uint64_t mode,
+	uint64_t uid, uint64_t gid, uint64_t rdev, uint64_t nodeid)
+{
+	bzero(vap, sizeof (*vap));
+	vap->va_mask = (uint_t)mask;
+	vap->va_type = IFTOVT(mode);
+	vap->va_mode = mode & MODEMASK;
+	vap->va_uid = (uid_t)(IS_EPHEMERAL(uid)) ? -1 : uid;
+	vap->va_gid = (gid_t)(IS_EPHEMERAL(gid)) ? -1 : gid;
+	vap->va_rdev = zfs_cmpldev(rdev);
+	vap->va_nodeid = nodeid;
+}
+
+/* ARGSUSED */
+static int
+zfs_replay_error(zfsvfs_t *zfsvfs, lr_t *lr, boolean_t byteswap)
+{
+	return (ENOTSUP);
+}
+
+static void
+zfs_replay_xvattr(lr_attr_t *lrattr, xvattr_t *xvap)
+{
+	xoptattr_t *xoap = NULL;
+	uint64_t *attrs;
+	uint64_t *crtime;
+	uint32_t *bitmap;
+	void *scanstamp;
+	int i;
+
+	xvap->xva_vattr.va_mask |= AT_XVATTR;
+	if ((xoap = xva_getxoptattr(xvap)) == NULL) {
+		xvap->xva_vattr.va_mask &= ~AT_XVATTR; /* shouldn't happen */
+		return;
+	}
+
+	ASSERT(lrattr->lr_attr_masksize == xvap->xva_mapsize);
+
+	bitmap = &lrattr->lr_attr_bitmap;
+	for (i = 0; i != lrattr->lr_attr_masksize; i++, bitmap++)
+		xvap->xva_reqattrmap[i] = *bitmap;
+
+	attrs = (uint64_t *)(lrattr + lrattr->lr_attr_masksize - 1);
+	crtime = attrs + 1;
+	scanstamp = (caddr_t)(crtime + 2);
+
+	if (XVA_ISSET_REQ(xvap, XAT_HIDDEN))
+		xoap->xoa_hidden = ((*attrs & XAT0_HIDDEN) != 0);
+	if (XVA_ISSET_REQ(xvap, XAT_SYSTEM))
+		xoap->xoa_system = ((*attrs & XAT0_SYSTEM) != 0);
+	if (XVA_ISSET_REQ(xvap, XAT_ARCHIVE))
+		xoap->xoa_archive = ((*attrs & XAT0_ARCHIVE) != 0);
+	if (XVA_ISSET_REQ(xvap, XAT_READONLY))
+		xoap->xoa_readonly = ((*attrs & XAT0_READONLY) != 0);
+	if (XVA_ISSET_REQ(xvap, XAT_IMMUTABLE))
+		xoap->xoa_immutable = ((*attrs & XAT0_IMMUTABLE) != 0);
+	if (XVA_ISSET_REQ(xvap, XAT_NOUNLINK))
+		xoap->xoa_nounlink = ((*attrs & XAT0_NOUNLINK) != 0);
+	if (XVA_ISSET_REQ(xvap, XAT_APPENDONLY))
+		xoap->xoa_appendonly = ((*attrs & XAT0_APPENDONLY) != 0);
+	if (XVA_ISSET_REQ(xvap, XAT_NODUMP))
+		xoap->xoa_nodump = ((*attrs & XAT0_NODUMP) != 0);
+	if (XVA_ISSET_REQ(xvap, XAT_OPAQUE))
+		xoap->xoa_opaque = ((*attrs & XAT0_OPAQUE) != 0);
+	if (XVA_ISSET_REQ(xvap, XAT_AV_MODIFIED))
+		xoap->xoa_av_modified = ((*attrs & XAT0_AV_MODIFIED) != 0);
+	if (XVA_ISSET_REQ(xvap, XAT_AV_QUARANTINED))
+		xoap->xoa_av_quarantined =
+		    ((*attrs & XAT0_AV_QUARANTINED) != 0);
+	if (XVA_ISSET_REQ(xvap, XAT_CREATETIME))
+		ZFS_TIME_DECODE(&xoap->xoa_createtime, crtime);
+	if (XVA_ISSET_REQ(xvap, XAT_AV_SCANSTAMP))
+		bcopy(scanstamp, xoap->xoa_av_scanstamp, AV_SCANSTAMP_SZ);
+}
+
+static int
+zfs_replay_domain_cnt(uint64_t uid, uint64_t gid)
+{
+	uint64_t uid_idx;
+	uint64_t gid_idx;
+	int domcnt = 0;
+
+	uid_idx = FUID_INDEX(uid);
+	gid_idx = FUID_INDEX(gid);
+	if (uid_idx)
+		domcnt++;
+	if (gid_idx > 0 && gid_idx != uid_idx)
+		domcnt++;
+
+	return (domcnt);
+}
+
+static void *
+zfs_replay_fuid_domain_common(zfs_fuid_info_t *fuid_infop, void *start,
+    int domcnt)
+{
+	int i;
+
+	for (i = 0; i != domcnt; i++) {
+		fuid_infop->z_domain_table[i] = start;
+		start = (caddr_t)start + strlen(start) + 1;
+	}
+
+	return (start);
+}
+
+/*
+ * Set the uid/gid in the fuid_info structure.
+ */
+static void
+zfs_replay_fuid_ugid(zfs_fuid_info_t *fuid_infop, uint64_t uid, uint64_t gid)
+{
+	/*
+	 * If owner or group are log specific FUIDs then slurp up
+	 * domain information and build zfs_fuid_info_t
+	 */
+	if (IS_EPHEMERAL(uid))
+		fuid_infop->z_fuid_owner = uid;
+
+	if (IS_EPHEMERAL(gid))
+		fuid_infop->z_fuid_group = gid;
+}
+
+/*
+ * Load fuid domains into fuid_info_t
+ */
+static zfs_fuid_info_t *
+zfs_replay_fuid_domain(void *buf, void **end, uint64_t uid, uint64_t gid)
+{
+	int domcnt;
+
+	zfs_fuid_info_t *fuid_infop;
+
+	fuid_infop = zfs_fuid_info_alloc();
+
+	domcnt = zfs_replay_domain_cnt(uid, gid);
+
+	if (domcnt == 0)
+		return (fuid_infop);
+
+	fuid_infop->z_domain_table =
+	    kmem_zalloc(domcnt * sizeof (char **), KM_SLEEP);
+
+	zfs_replay_fuid_ugid(fuid_infop, uid, gid);
+
+	fuid_infop->z_domain_cnt = domcnt;
+	*end = zfs_replay_fuid_domain_common(fuid_infop, buf, domcnt);
+	return (fuid_infop);
+}
+
+/*
+ * load zfs_fuid_t's and fuid_domains into fuid_info_t
+ */
+static zfs_fuid_info_t *
+zfs_replay_fuids(void *start, void **end, int idcnt, int domcnt, uint64_t uid,
+    uint64_t gid)
+{
+	uint64_t *log_fuid = (uint64_t *)start;
+	zfs_fuid_info_t *fuid_infop;
+	int i;
+
+	fuid_infop = zfs_fuid_info_alloc();
+	fuid_infop->z_domain_cnt = domcnt;
+
+	fuid_infop->z_domain_table =
+	    kmem_zalloc(domcnt * sizeof (char **), KM_SLEEP);
+
+	for (i = 0; i != idcnt; i++) {
+		zfs_fuid_t *zfuid;
+
+		zfuid = kmem_alloc(sizeof (zfs_fuid_t), KM_SLEEP);
+		zfuid->z_logfuid = *log_fuid;
+		zfuid->z_id = -1;
+		zfuid->z_domidx = 0;
+		list_insert_tail(&fuid_infop->z_fuids, zfuid);
+		log_fuid++;
+	}
+
+	zfs_replay_fuid_ugid(fuid_infop, uid, gid);
+
+	*end = zfs_replay_fuid_domain_common(fuid_infop, log_fuid, domcnt);
+	return (fuid_infop);
+}
+
+static void
+zfs_replay_swap_attrs(lr_attr_t *lrattr)
+{
+	/* swap the lr_attr structure */
+	byteswap_uint32_array(lrattr, sizeof (*lrattr));
+	/* swap the bitmap */
+	byteswap_uint32_array(lrattr + 1, (lrattr->lr_attr_masksize - 1) *
+	    sizeof (uint32_t));
+	/* swap the attributes, create time + 64 bit word for attributes */
+	byteswap_uint64_array((caddr_t)(lrattr + 1) + (sizeof (uint32_t) *
+	    (lrattr->lr_attr_masksize - 1)), 3 * sizeof (uint64_t));
+}
+
+/*
+ * Replay file create with optional ACL, xvattr information as well
+ * as option FUID information.
+ */
+static int
+zfs_replay_create_acl(zfsvfs_t *zfsvfs,
+    lr_acl_create_t *lracl, boolean_t byteswap)
+{
+	char *name = NULL;		/* location determined later */
+	lr_create_t *lr = (lr_create_t *)lracl;
+	znode_t *dzp;
+	vnode_t *vp = NULL;
+	xvattr_t xva;
+	int vflg = 0;
+	vsecattr_t vsec = { 0 };
+	lr_attr_t *lrattr;
+	void *aclstart;
+	void *fuidstart;
+	size_t xvatlen = 0;
+	uint64_t txtype;
+	int error;
+
+	if (byteswap) {
+		byteswap_uint64_array(lracl, sizeof (*lracl));
+		txtype = (int)lr->lr_common.lrc_txtype;
+		if (txtype == TX_CREATE_ACL_ATTR ||
+		    txtype == TX_MKDIR_ACL_ATTR) {
+			lrattr = (lr_attr_t *)(caddr_t)(lracl + 1);
+			zfs_replay_swap_attrs(lrattr);
+			xvatlen = ZIL_XVAT_SIZE(lrattr->lr_attr_masksize);
+		}
+
+		aclstart = (caddr_t)(lracl + 1) + xvatlen;
+		zfs_ace_byteswap(aclstart, lracl->lr_acl_bytes, B_FALSE);
+		/* swap fuids */
+		if (lracl->lr_fuidcnt) {
+			byteswap_uint64_array((caddr_t)aclstart +
+			    ZIL_ACE_LENGTH(lracl->lr_acl_bytes),
+			    lracl->lr_fuidcnt * sizeof (uint64_t));
+		}
+	}
+
+	if ((error = zfs_zget(zfsvfs, lr->lr_doid, &dzp)) != 0)
+		return (error);
+
+	xva_init(&xva);
+	zfs_init_vattr(&xva.xva_vattr, AT_TYPE | AT_MODE | AT_UID | AT_GID,
+	    lr->lr_mode, lr->lr_uid, lr->lr_gid, lr->lr_rdev, lr->lr_foid);
+
+	/*
+	 * All forms of zfs create (create, mkdir, mkxattrdir, symlink)
+	 * eventually end up in zfs_mknode(), which assigns the object's
+	 * creation time and generation number.  The generic VOP_CREATE()
+	 * doesn't have either concept, so we smuggle the values inside
+	 * the vattr's otherwise unused va_ctime and va_nblocks fields.
+	 */
+	ZFS_TIME_DECODE(&xva.xva_vattr.va_ctime, lr->lr_crtime);
+	xva.xva_vattr.va_nblocks = lr->lr_gen;
+
+	error = dmu_object_info(zfsvfs->z_os, lr->lr_foid, NULL);
+	if (error != ENOENT)
+		goto bail;
+
+	if (lr->lr_common.lrc_txtype & TX_CI)
+		vflg |= FIGNORECASE;
+	switch ((int)lr->lr_common.lrc_txtype) {
+	case TX_CREATE_ACL:
+		aclstart = (caddr_t)(lracl + 1);
+		fuidstart = (caddr_t)aclstart +
+		    ZIL_ACE_LENGTH(lracl->lr_acl_bytes);
+		zfsvfs->z_fuid_replay = zfs_replay_fuids(fuidstart,
+		    (void *)&name, lracl->lr_fuidcnt, lracl->lr_domcnt,
+		    lr->lr_uid, lr->lr_gid);
+		/*FALLTHROUGH*/
+	case TX_CREATE_ACL_ATTR:
+		if (name == NULL) {
+			lrattr = (lr_attr_t *)(caddr_t)(lracl + 1);
+			xvatlen = ZIL_XVAT_SIZE(lrattr->lr_attr_masksize);
+			xva.xva_vattr.va_mask |= AT_XVATTR;
+			zfs_replay_xvattr(lrattr, &xva);
+		}
+		vsec.vsa_mask = VSA_ACE | VSA_ACE_ACLFLAGS;
+		vsec.vsa_aclentp = (caddr_t)(lracl + 1) + xvatlen;
+		vsec.vsa_aclcnt = lracl->lr_aclcnt;
+		vsec.vsa_aclentsz = lracl->lr_acl_bytes;
+		vsec.vsa_aclflags = lracl->lr_acl_flags;
+		if (zfsvfs->z_fuid_replay == NULL) {
+			fuidstart = (caddr_t)(lracl + 1) + xvatlen +
+			    ZIL_ACE_LENGTH(lracl->lr_acl_bytes);
+			zfsvfs->z_fuid_replay =
+			    zfs_replay_fuids(fuidstart,
+			    (void *)&name, lracl->lr_fuidcnt, lracl->lr_domcnt,
+			    lr->lr_uid, lr->lr_gid);
+		}
+
+		error = VOP_CREATE(ZTOV(dzp), name, &xva.xva_vattr,
+		    0, 0, &vp, kcred, vflg, NULL, &vsec);
+		break;
+	case TX_MKDIR_ACL:
+		aclstart = (caddr_t)(lracl + 1);
+		fuidstart = (caddr_t)aclstart +
+		    ZIL_ACE_LENGTH(lracl->lr_acl_bytes);
+		zfsvfs->z_fuid_replay = zfs_replay_fuids(fuidstart,
+		    (void *)&name, lracl->lr_fuidcnt, lracl->lr_domcnt,
+		    lr->lr_uid, lr->lr_gid);
+		/*FALLTHROUGH*/
+	case TX_MKDIR_ACL_ATTR:
+		if (name == NULL) {
+			lrattr = (lr_attr_t *)(caddr_t)(lracl + 1);
+			xvatlen = ZIL_XVAT_SIZE(lrattr->lr_attr_masksize);
+			zfs_replay_xvattr(lrattr, &xva);
+		}
+		vsec.vsa_mask = VSA_ACE | VSA_ACE_ACLFLAGS;
+		vsec.vsa_aclentp = (caddr_t)(lracl + 1) + xvatlen;
+		vsec.vsa_aclcnt = lracl->lr_aclcnt;
+		vsec.vsa_aclentsz = lracl->lr_acl_bytes;
+		vsec.vsa_aclflags = lracl->lr_acl_flags;
+		if (zfsvfs->z_fuid_replay == NULL) {
+			fuidstart = (caddr_t)(lracl + 1) + xvatlen +
+			    ZIL_ACE_LENGTH(lracl->lr_acl_bytes);
+			zfsvfs->z_fuid_replay =
+			    zfs_replay_fuids(fuidstart,
+			    (void *)&name, lracl->lr_fuidcnt, lracl->lr_domcnt,
+			    lr->lr_uid, lr->lr_gid);
+		}
+		error = VOP_MKDIR(ZTOV(dzp), name, &xva.xva_vattr,
+		    &vp, kcred, NULL, vflg, &vsec);
+		break;
+	default:
+		error = ENOTSUP;
+	}
+
+bail:
+	if (error == 0 && vp != NULL)
+		VN_RELE(vp);
+
+	VN_RELE(ZTOV(dzp));
+
+	zfs_fuid_info_free(zfsvfs->z_fuid_replay);
+	zfsvfs->z_fuid_replay = NULL;
+
+	return (error);
+}
+
+static int
+zfs_replay_create(zfsvfs_t *zfsvfs, lr_create_t *lr, boolean_t byteswap)
+{
+	char *name = NULL;		/* location determined later */
+	char *link;			/* symlink content follows name */
+	znode_t *dzp;
+	vnode_t *vp = NULL;
+	xvattr_t xva;
+	int vflg = 0;
+	size_t lrsize = sizeof (lr_create_t);
+	lr_attr_t *lrattr;
+	void *start;
+	size_t xvatlen;
+	uint64_t txtype;
+	int error;
+
+	if (byteswap) {
+		byteswap_uint64_array(lr, sizeof (*lr));
+		txtype = (int)lr->lr_common.lrc_txtype;
+		if (txtype == TX_CREATE_ATTR || txtype == TX_MKDIR_ATTR)
+			zfs_replay_swap_attrs((lr_attr_t *)(lr + 1));
+	}
+
+
+	if ((error = zfs_zget(zfsvfs, lr->lr_doid, &dzp)) != 0)
+		return (error);
+
+	xva_init(&xva);
+	zfs_init_vattr(&xva.xva_vattr, AT_TYPE | AT_MODE | AT_UID | AT_GID,
+	    lr->lr_mode, lr->lr_uid, lr->lr_gid, lr->lr_rdev, lr->lr_foid);
+
+	/*
+	 * All forms of zfs create (create, mkdir, mkxattrdir, symlink)
+	 * eventually end up in zfs_mknode(), which assigns the object's
+	 * creation time and generation number.  The generic VOP_CREATE()
+	 * doesn't have either concept, so we smuggle the values inside
+	 * the vattr's otherwise unused va_ctime and va_nblocks fields.
+	 */
+	ZFS_TIME_DECODE(&xva.xva_vattr.va_ctime, lr->lr_crtime);
+	xva.xva_vattr.va_nblocks = lr->lr_gen;
+
+	error = dmu_object_info(zfsvfs->z_os, lr->lr_foid, NULL);
+	if (error != ENOENT)
+		goto out;
+
+	if (lr->lr_common.lrc_txtype & TX_CI)
+		vflg |= FIGNORECASE;
+
+	/*
+	 * Symlinks don't have fuid info, and CIFS never creates
+	 * symlinks.
+	 *
+	 * The _ATTR versions will grab the fuid info in their subcases.
+	 */
+	if ((int)lr->lr_common.lrc_txtype != TX_SYMLINK &&
+	    (int)lr->lr_common.lrc_txtype != TX_MKDIR_ATTR &&
+	    (int)lr->lr_common.lrc_txtype != TX_CREATE_ATTR) {
+		start = (lr + 1);
+		zfsvfs->z_fuid_replay =
+		    zfs_replay_fuid_domain(start, &start,
+		    lr->lr_uid, lr->lr_gid);
+	}
+
+	switch ((int)lr->lr_common.lrc_txtype) {
+	case TX_CREATE_ATTR:
+		lrattr = (lr_attr_t *)(caddr_t)(lr + 1);
+		xvatlen = ZIL_XVAT_SIZE(lrattr->lr_attr_masksize);
+		zfs_replay_xvattr((lr_attr_t *)((caddr_t)lr + lrsize), &xva);
+		start = (caddr_t)(lr + 1) + xvatlen;
+		zfsvfs->z_fuid_replay =
+		    zfs_replay_fuid_domain(start, &start,
+		    lr->lr_uid, lr->lr_gid);
+		name = (char *)start;
+
+		/*FALLTHROUGH*/
+	case TX_CREATE:
+		if (name == NULL)
+			name = (char *)start;
+
+		error = VOP_CREATE(ZTOV(dzp), name, &xva.xva_vattr,
+		    0, 0, &vp, kcred, vflg, NULL, NULL);
+		break;
+	case TX_MKDIR_ATTR:
+		lrattr = (lr_attr_t *)(caddr_t)(lr + 1);
+		xvatlen = ZIL_XVAT_SIZE(lrattr->lr_attr_masksize);
+		zfs_replay_xvattr((lr_attr_t *)((caddr_t)lr + lrsize), &xva);
+		start = (caddr_t)(lr + 1) + xvatlen;
+		zfsvfs->z_fuid_replay =
+		    zfs_replay_fuid_domain(start, &start,
+		    lr->lr_uid, lr->lr_gid);
+		name = (char *)start;
+
+		/*FALLTHROUGH*/
+	case TX_MKDIR:
+		if (name == NULL)
+			name = (char *)(lr + 1);
+
+		error = VOP_MKDIR(ZTOV(dzp), name, &xva.xva_vattr,
+		    &vp, kcred, NULL, vflg, NULL);
+		break;
+	case TX_MKXATTR:
+		name = (char *)(lr + 1);
+		error = zfs_make_xattrdir(dzp, &xva.xva_vattr, &vp, kcred);
+		break;
+	case TX_SYMLINK:
+		name = (char *)(lr + 1);
+		link = name + strlen(name) + 1;
+		error = VOP_SYMLINK(ZTOV(dzp), name, &xva.xva_vattr,
+		    link, kcred, NULL, vflg);
+		break;
+	default:
+		error = ENOTSUP;
+	}
+
+out:
+	if (error == 0 && vp != NULL)
+		VN_RELE(vp);
+
+	VN_RELE(ZTOV(dzp));
+
+	if (zfsvfs->z_fuid_replay)
+		zfs_fuid_info_free(zfsvfs->z_fuid_replay);
+	zfsvfs->z_fuid_replay = NULL;
+	return (error);
+}
+
+static int
+zfs_replay_remove(zfsvfs_t *zfsvfs, lr_remove_t *lr, boolean_t byteswap)
+{
+	char *name = (char *)(lr + 1);	/* name follows lr_remove_t */
+	znode_t *dzp;
+	int error;
+	int vflg = 0;
+
+	if (byteswap)
+		byteswap_uint64_array(lr, sizeof (*lr));
+
+	if ((error = zfs_zget(zfsvfs, lr->lr_doid, &dzp)) != 0)
+		return (error);
+
+	if (lr->lr_common.lrc_txtype & TX_CI)
+		vflg |= FIGNORECASE;
+
+	switch ((int)lr->lr_common.lrc_txtype) {
+	case TX_REMOVE:
+		error = VOP_REMOVE(ZTOV(dzp), name, kcred, NULL, vflg);
+		break;
+	case TX_RMDIR:
+		error = VOP_RMDIR(ZTOV(dzp), name, NULL, kcred, NULL, vflg);
+		break;
+	default:
+		error = ENOTSUP;
+	}
+
+	VN_RELE(ZTOV(dzp));
+
+	return (error);
+}
+
+static int
+zfs_replay_link(zfsvfs_t *zfsvfs, lr_link_t *lr, boolean_t byteswap)
+{
+	char *name = (char *)(lr + 1);	/* name follows lr_link_t */
+	znode_t *dzp, *zp;
+	int error;
+	int vflg = 0;
+
+	if (byteswap)
+		byteswap_uint64_array(lr, sizeof (*lr));
+
+	if ((error = zfs_zget(zfsvfs, lr->lr_doid, &dzp)) != 0)
+		return (error);
+
+	if ((error = zfs_zget(zfsvfs, lr->lr_link_obj, &zp)) != 0) {
+		VN_RELE(ZTOV(dzp));
+		return (error);
+	}
+
+	if (lr->lr_common.lrc_txtype & TX_CI)
+		vflg |= FIGNORECASE;
+
+	error = VOP_LINK(ZTOV(dzp), ZTOV(zp), name, kcred, NULL, vflg);
+
+	VN_RELE(ZTOV(zp));
+	VN_RELE(ZTOV(dzp));
+
+	return (error);
+}
+
+static int
+zfs_replay_rename(zfsvfs_t *zfsvfs, lr_rename_t *lr, boolean_t byteswap)
+{
+	char *sname = (char *)(lr + 1);	/* sname and tname follow lr_rename_t */
+	char *tname = sname + strlen(sname) + 1;
+	znode_t *sdzp, *tdzp;
+	int error;
+	int vflg = 0;
+
+	if (byteswap)
+		byteswap_uint64_array(lr, sizeof (*lr));
+
+	if ((error = zfs_zget(zfsvfs, lr->lr_sdoid, &sdzp)) != 0)
+		return (error);
+
+	if ((error = zfs_zget(zfsvfs, lr->lr_tdoid, &tdzp)) != 0) {
+		VN_RELE(ZTOV(sdzp));
+		return (error);
+	}
+
+	if (lr->lr_common.lrc_txtype & TX_CI)
+		vflg |= FIGNORECASE;
+
+	error = VOP_RENAME(ZTOV(sdzp), sname, ZTOV(tdzp), tname, kcred,
+	    NULL, vflg);
+
+	VN_RELE(ZTOV(tdzp));
+	VN_RELE(ZTOV(sdzp));
+
+	return (error);
+}
+
+static int
+zfs_replay_write(zfsvfs_t *zfsvfs, lr_write_t *lr, boolean_t byteswap)
+{
+	char *data = (char *)(lr + 1);	/* data follows lr_write_t */
+	znode_t	*zp;
+	int error;
+	ssize_t resid;
+
+	if (byteswap)
+		byteswap_uint64_array(lr, sizeof (*lr));
+
+	if ((error = zfs_zget(zfsvfs, lr->lr_foid, &zp)) != 0) {
+		/*
+		 * As we can log writes out of order, it's possible the
+		 * file has been removed. In this case just drop the write
+		 * and return success.
+		 */
+		if (error == ENOENT)
+			error = 0;
+		return (error);
+	}
+
+	error = vn_rdwr(UIO_WRITE, ZTOV(zp), data, lr->lr_length,
+	    lr->lr_offset, UIO_SYSSPACE, 0, RLIM64_INFINITY, kcred, &resid);
+
+	VN_RELE(ZTOV(zp));
+
+	return (error);
+}
+
+static int
+zfs_replay_truncate(zfsvfs_t *zfsvfs, lr_truncate_t *lr, boolean_t byteswap)
+{
+	znode_t *zp;
+	flock64_t fl;
+	int error;
+
+	if (byteswap)
+		byteswap_uint64_array(lr, sizeof (*lr));
+
+	if ((error = zfs_zget(zfsvfs, lr->lr_foid, &zp)) != 0) {
+		/*
+		 * As we can log truncates out of order, it's possible the
+		 * file has been removed. In this case just drop the truncate
+		 * and return success.
+		 */
+		if (error == ENOENT)
+			error = 0;
+		return (error);
+	}
+
+	bzero(&fl, sizeof (fl));
+	fl.l_type = F_WRLCK;
+	fl.l_whence = 0;
+	fl.l_start = lr->lr_offset;
+	fl.l_len = lr->lr_length;
+
+	error = VOP_SPACE(ZTOV(zp), F_FREESP, &fl, FWRITE | FOFFMAX,
+	    lr->lr_offset, kcred, NULL);
+
+	VN_RELE(ZTOV(zp));
+
+	return (error);
+}
+
+static int
+zfs_replay_setattr(zfsvfs_t *zfsvfs, lr_setattr_t *lr, boolean_t byteswap)
+{
+	znode_t *zp;
+	xvattr_t xva;
+	vattr_t *vap = &xva.xva_vattr;
+	int error;
+	void *start;
+
+	xva_init(&xva);
+	if (byteswap) {
+		byteswap_uint64_array(lr, sizeof (*lr));
+
+		if ((lr->lr_mask & AT_XVATTR) &&
+		    zfsvfs->z_version >= ZPL_VERSION_INITIAL)
+			zfs_replay_swap_attrs((lr_attr_t *)(lr + 1));
+	}
+
+	if ((error = zfs_zget(zfsvfs, lr->lr_foid, &zp)) != 0) {
+		/*
+		 * As we can log setattrs out of order, it's possible the
+		 * file has been removed. In this case just drop the setattr
+		 * and return success.
+		 */
+		if (error == ENOENT)
+			error = 0;
+		return (error);
+	}
+
+	zfs_init_vattr(vap, lr->lr_mask, lr->lr_mode,
+	    lr->lr_uid, lr->lr_gid, 0, lr->lr_foid);
+
+	vap->va_size = lr->lr_size;
+	ZFS_TIME_DECODE(&vap->va_atime, lr->lr_atime);
+	ZFS_TIME_DECODE(&vap->va_mtime, lr->lr_mtime);
+
+	/*
+	 * Fill in xvattr_t portions if necessary.
+	 */
+
+	start = (lr_setattr_t *)(lr + 1);
+	if (vap->va_mask & AT_XVATTR) {
+		zfs_replay_xvattr((lr_attr_t *)start, &xva);
+		start = (caddr_t)start +
+		    ZIL_XVAT_SIZE(((lr_attr_t *)start)->lr_attr_masksize);
+	} else
+		xva.xva_vattr.va_mask &= ~AT_XVATTR;
+
+	zfsvfs->z_fuid_replay = zfs_replay_fuid_domain(start, &start,
+	    lr->lr_uid, lr->lr_gid);
+
+	error = VOP_SETATTR(ZTOV(zp), vap, 0, kcred, NULL);
+
+	zfs_fuid_info_free(zfsvfs->z_fuid_replay);
+	zfsvfs->z_fuid_replay = NULL;
+	VN_RELE(ZTOV(zp));
+
+	return (error);
+}
+
+static int
+zfs_replay_acl_v0(zfsvfs_t *zfsvfs, lr_acl_v0_t *lr, boolean_t byteswap)
+{
+	ace_t *ace = (ace_t *)(lr + 1);	/* ace array follows lr_acl_t */
+	vsecattr_t vsa;
+	znode_t *zp;
+	int error;
+
+	if (byteswap) {
+		byteswap_uint64_array(lr, sizeof (*lr));
+		zfs_oldace_byteswap(ace, lr->lr_aclcnt);
+	}
+
+	if ((error = zfs_zget(zfsvfs, lr->lr_foid, &zp)) != 0) {
+		/*
+		 * As we can log acls out of order, it's possible the
+		 * file has been removed. In this case just drop the acl
+		 * and return success.
+		 */
+		if (error == ENOENT)
+			error = 0;
+		return (error);
+	}
+
+	bzero(&vsa, sizeof (vsa));
+	vsa.vsa_mask = VSA_ACE | VSA_ACECNT;
+	vsa.vsa_aclcnt = lr->lr_aclcnt;
+	vsa.vsa_aclentp = ace;
+
+	error = VOP_SETSECATTR(ZTOV(zp), &vsa, 0, kcred, NULL);
+
+	VN_RELE(ZTOV(zp));
+
+	return (error);
+}
+
+/*
+ * Replaying ACLs is complicated by FUID support.
+ * The log record may contain some optional data
+ * to be used for replaying FUID's.  These pieces
+ * are the actual FUIDs that were created initially.
+ * The FUID table index may no longer be valid and
+ * during zfs_create() a new index may be assigned.
+ * Because of this the log will contain the original
+ * doman+rid in order to create a new FUID.
+ *
+ * The individual ACEs may contain an ephemeral uid/gid which is no
+ * longer valid and will need to be replaced with an actual FUID.
+ *
+ */
+static int
+zfs_replay_acl(zfsvfs_t *zfsvfs, lr_acl_t *lr, boolean_t byteswap)
+{
+	ace_t *ace = (ace_t *)(lr + 1);
+	vsecattr_t vsa;
+	znode_t *zp;
+	int error;
+
+	if (byteswap) {
+		byteswap_uint64_array(lr, sizeof (*lr));
+		zfs_ace_byteswap(ace, lr->lr_acl_bytes, B_FALSE);
+		if (lr->lr_fuidcnt) {
+			byteswap_uint64_array((caddr_t)ace +
+			    ZIL_ACE_LENGTH(lr->lr_acl_bytes),
+			    lr->lr_fuidcnt * sizeof (uint64_t));
+		}
+	}
+
+	if ((error = zfs_zget(zfsvfs, lr->lr_foid, &zp)) != 0) {
+		/*
+		 * As we can log acls out of order, it's possible the
+		 * file has been removed. In this case just drop the acl
+		 * and return success.
+		 */
+		if (error == ENOENT)
+			error = 0;
+		return (error);
+	}
+
+	bzero(&vsa, sizeof (vsa));
+	vsa.vsa_mask = VSA_ACE | VSA_ACECNT | VSA_ACE_ACLFLAGS;
+	vsa.vsa_aclcnt = lr->lr_aclcnt;
+	vsa.vsa_aclentp = ace;
+	vsa.vsa_aclentsz = lr->lr_acl_bytes;
+	vsa.vsa_aclflags = lr->lr_acl_flags;
+
+	if (lr->lr_fuidcnt) {
+		void *fuidstart = (caddr_t)ace +
+		    ZIL_ACE_LENGTH(lr->lr_acl_bytes);
+
+		zfsvfs->z_fuid_replay =
+		    zfs_replay_fuids(fuidstart, &fuidstart,
+		    lr->lr_fuidcnt, lr->lr_domcnt, 0, 0);
+	}
+
+	error = VOP_SETSECATTR(ZTOV(zp), &vsa, 0, kcred, NULL);
+
+	if (zfsvfs->z_fuid_replay)
+		zfs_fuid_info_free(zfsvfs->z_fuid_replay);
+
+	zfsvfs->z_fuid_replay = NULL;
+	VN_RELE(ZTOV(zp));
+
+	return (error);
+}
+
+/*
+ * Callback vectors for replaying records
+ */
+zil_replay_func_t *zfs_replay_vector[TX_MAX_TYPE] = {
+	zfs_replay_error,	/* 0 no such transaction type */
+	zfs_replay_create,	/* TX_CREATE */
+	zfs_replay_create,	/* TX_MKDIR */
+	zfs_replay_create,	/* TX_MKXATTR */
+	zfs_replay_create,	/* TX_SYMLINK */
+	zfs_replay_remove,	/* TX_REMOVE */
+	zfs_replay_remove,	/* TX_RMDIR */
+	zfs_replay_link,	/* TX_LINK */
+	zfs_replay_rename,	/* TX_RENAME */
+	zfs_replay_write,	/* TX_WRITE */
+	zfs_replay_truncate,	/* TX_TRUNCATE */
+	zfs_replay_setattr,	/* TX_SETATTR */
+	zfs_replay_acl_v0,	/* TX_ACL_V0 */
+	zfs_replay_acl,		/* TX_ACL */
+	zfs_replay_create_acl,	/* TX_CREATE_ACL */
+	zfs_replay_create,	/* TX_CREATE_ATTR */
+	zfs_replay_create_acl,	/* TX_CREATE_ACL_ATTR */
+	zfs_replay_create_acl,	/* TX_MKDIR_ACL */
+	zfs_replay_create,	/* TX_MKDIR_ATTR */
+	zfs_replay_create_acl,	/* TX_MKDIR_ACL_ATTR */
+};
--- a/zfs/lib/libdmu-ctl/zfs_rlock.c
+++ b/zfs/lib/libdmu-ctl/zfs_rlock.c
@ -0,0 +1,602 @@
+/*
+ * CDDL HEADER START
+ *
+ * The contents of this file are subject to the terms of the
+ * Common Development and Distribution License (the "License").
+ * You may not use this file except in compliance with the License.
+ *
+ * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
+ * or http://www.opensolaris.org/os/licensing.
+ * See the License for the specific language governing permissions
+ * and limitations under the License.
+ *
+ * When distributing Covered Code, include this CDDL HEADER in each
+ * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
+ * If applicable, add the following below this CDDL HEADER, with the
+ * fields enclosed by brackets "[]" replaced with your own identifying
+ * information: Portions Copyright [yyyy] [name of copyright owner]
+ *
+ * CDDL HEADER END
+ */
+/*
+ * Copyright 2007 Sun Microsystems, Inc.  All rights reserved.
+ * Use is subject to license terms.
+ */
+
+#pragma ident	"@(#)zfs_rlock.c	1.4	07/08/08 SMI"
+
+/*
+ * This file contains the code to implement file range locking in
+ * ZFS, although there isn't much specific to ZFS (all that comes to mind
+ * support for growing the blocksize).
+ *
+ * Interface
+ * ---------
+ * Defined in zfs_rlock.h but essentially:
+ *	rl = zfs_range_lock(zp, off, len, lock_type);
+ *	zfs_range_unlock(rl);
+ *	zfs_range_reduce(rl, off, len);
+ *
+ * AVL tree
+ * --------
+ * An AVL tree is used to maintain the state of the existing ranges
+ * that are locked for exclusive (writer) or shared (reader) use.
+ * The starting range offset is used for searching and sorting the tree.
+ *
+ * Common case
+ * -----------
+ * The (hopefully) usual case is of no overlaps or contention for
+ * locks. On entry to zfs_lock_range() a rl_t is allocated; the tree
+ * searched that finds no overlap, and *this* rl_t is placed in the tree.
+ *
+ * Overlaps/Reference counting/Proxy locks
+ * ---------------------------------------
+ * The avl code only allows one node at a particular offset. Also it's very
+ * inefficient to search through all previous entries looking for overlaps
+ * (because the very 1st in the ordered list might be at offset 0 but
+ * cover the whole file).
+ * So this implementation uses reference counts and proxy range locks.
+ * Firstly, only reader locks use reference counts and proxy locks,
+ * because writer locks are exclusive.
+ * When a reader lock overlaps with another then a proxy lock is created
+ * for that range and replaces the original lock. If the overlap
+ * is exact then the reference count of the proxy is simply incremented.
+ * Otherwise, the proxy lock is split into smaller lock ranges and
+ * new proxy locks created for non overlapping ranges.
+ * The reference counts are adjusted accordingly.
+ * Meanwhile, the orginal lock is kept around (this is the callers handle)
+ * and its offset and length are used when releasing the lock.
+ *
+ * Thread coordination
+ * -------------------
+ * In order to make wakeups efficient and to ensure multiple continuous
+ * readers on a range don't starve a writer for the same range lock,
+ * two condition variables are allocated in each rl_t.
+ * If a writer (or reader) can't get a range it initialises the writer
+ * (or reader) cv; sets a flag saying there's a writer (or reader) waiting;
+ * and waits on that cv. When a thread unlocks that range it wakes up all
+ * writers then all readers before destroying the lock.
+ *
+ * Append mode writes
+ * ------------------
+ * Append mode writes need to lock a range at the end of a file.
+ * The offset of the end of the file is determined under the
+ * range locking mutex, and the lock type converted from RL_APPEND to
+ * RL_WRITER and the range locked.
+ *
+ * Grow block handling
+ * -------------------
+ * ZFS supports multiple block sizes currently upto 128K. The smallest
+ * block size is used for the file which is grown as needed. During this
+ * growth all other writers and readers must be excluded.
+ * So if the block size needs to be grown then the whole file is
+ * exclusively locked, then later the caller will reduce the lock
+ * range to just the range to be written using zfs_reduce_range.
+ */
+
+#include <sys/zfs_rlock.h>
+
+/*
+ * Check if a write lock can be grabbed, or wait and recheck until available.
+ */
+static void
+zfs_range_lock_writer(znode_t *zp, rl_t *new)
+{
+	avl_tree_t *tree = &zp->z_range_avl;
+	rl_t *rl;
+	avl_index_t where;
+	uint64_t end_size;
+	uint64_t off = new->r_off;
+	uint64_t len = new->r_len;
+
+	for (;;) {
+		/*
+		 * Range locking is also used by zvol and uses a
+		 * dummied up znode. However, for zvol, we don't need to
+		 * append or grow blocksize, and besides we don't have
+		 * a z_phys or z_zfsvfs - so skip that processing.
+		 *
+		 * Yes, this is ugly, and would be solved by not handling
+		 * grow or append in range lock code. If that was done then
+		 * we could make the range locking code generically available
+		 * to other non-zfs consumers.
+		 */
+		if (zp->z_vnode) { /* caller is ZPL */
+			/*
+			 * If in append mode pick up the current end of file.
+			 * This is done under z_range_lock to avoid races.
+			 */
+			if (new->r_type == RL_APPEND)
+				new->r_off = zp->z_phys->zp_size;
+
+			/*
+			 * If we need to grow the block size then grab the whole
+			 * file range. This is also done under z_range_lock to
+			 * avoid races.
+			 */
+			end_size = MAX(zp->z_phys->zp_size, new->r_off + len);
+			if (end_size > zp->z_blksz && (!ISP2(zp->z_blksz) ||
+			    zp->z_blksz < zp->z_zfsvfs->z_max_blksz)) {
+				new->r_off = 0;
+				new->r_len = UINT64_MAX;
+			}
+		}
+
+		/*
+		 * First check for the usual case of no locks
+		 */
+		if (avl_numnodes(tree) == 0) {
+			new->r_type = RL_WRITER; /* convert to writer */
+			avl_add(tree, new);
+			return;
+		}
+
+		/*
+		 * Look for any locks in the range.
+		 */
+		rl = avl_find(tree, new, &where);
+		if (rl)
+			goto wait; /* already locked at same offset */
+
+		rl = (rl_t *)avl_nearest(tree, where, AVL_AFTER);
+		if (rl && (rl->r_off < new->r_off + new->r_len))
+			goto wait;
+
+		rl = (rl_t *)avl_nearest(tree, where, AVL_BEFORE);
+		if (rl && rl->r_off + rl->r_len > new->r_off)
+			goto wait;
+
+		new->r_type = RL_WRITER; /* convert possible RL_APPEND */
+		avl_insert(tree, new, where);
+		return;
+wait:
+		if (!rl->r_write_wanted) {
+			cv_init(&rl->r_wr_cv, NULL, CV_DEFAULT, NULL);
+			rl->r_write_wanted = B_TRUE;
+		}
+		cv_wait(&rl->r_wr_cv, &zp->z_range_lock);
+
+		/* reset to original */
+		new->r_off = off;
+		new->r_len = len;
+	}
+}
+
+/*
+ * If this is an original (non-proxy) lock then replace it by
+ * a proxy and return the proxy.
+ */
+static rl_t *
+zfs_range_proxify(avl_tree_t *tree, rl_t *rl)
+{
+	rl_t *proxy;
+
+	if (rl->r_proxy)
+		return (rl); /* already a proxy */
+
+	ASSERT3U(rl->r_cnt, ==, 1);
+	ASSERT(rl->r_write_wanted == B_FALSE);
+	ASSERT(rl->r_read_wanted == B_FALSE);
+	avl_remove(tree, rl);
+	rl->r_cnt = 0;
+
+	/* create a proxy range lock */
+	proxy = kmem_alloc(sizeof (rl_t), KM_SLEEP);
+	proxy->r_off = rl->r_off;
+	proxy->r_len = rl->r_len;
+	proxy->r_cnt = 1;
+	proxy->r_type = RL_READER;
+	proxy->r_proxy = B_TRUE;
+	proxy->r_write_wanted = B_FALSE;
+	proxy->r_read_wanted = B_FALSE;
+	avl_add(tree, proxy);
+
+	return (proxy);
+}
+
+/*
+ * Split the range lock at the supplied offset
+ * returning the *front* proxy.
+ */
+static rl_t *
+zfs_range_split(avl_tree_t *tree, rl_t *rl, uint64_t off)
+{
+	rl_t *front, *rear;
+
+	ASSERT3U(rl->r_len, >, 1);
+	ASSERT3U(off, >, rl->r_off);
+	ASSERT3U(off, <, rl->r_off + rl->r_len);
+	ASSERT(rl->r_write_wanted == B_FALSE);
+	ASSERT(rl->r_read_wanted == B_FALSE);
+
+	/* create the rear proxy range lock */
+	rear = kmem_alloc(sizeof (rl_t), KM_SLEEP);
+	rear->r_off = off;
+	rear->r_len = rl->r_off + rl->r_len - off;
+	rear->r_cnt = rl->r_cnt;
+	rear->r_type = RL_READER;
+	rear->r_proxy = B_TRUE;
+	rear->r_write_wanted = B_FALSE;
+	rear->r_read_wanted = B_FALSE;
+
+	front = zfs_range_proxify(tree, rl);
+	front->r_len = off - rl->r_off;
+
+	avl_insert_here(tree, rear, front, AVL_AFTER);
+	return (front);
+}
+
+/*
+ * Create and add a new proxy range lock for the supplied range.
+ */
+static void
+zfs_range_new_proxy(avl_tree_t *tree, uint64_t off, uint64_t len)
+{
+	rl_t *rl;
+
+	ASSERT(len);
+	rl = kmem_alloc(sizeof (rl_t), KM_SLEEP);
+	rl->r_off = off;
+	rl->r_len = len;
+	rl->r_cnt = 1;
+	rl->r_type = RL_READER;
+	rl->r_proxy = B_TRUE;
+	rl->r_write_wanted = B_FALSE;
+	rl->r_read_wanted = B_FALSE;
+	avl_add(tree, rl);
+}
+
+static void
+zfs_range_add_reader(avl_tree_t *tree, rl_t *new, rl_t *prev, avl_index_t where)
+{
+	rl_t *next;
+	uint64_t off = new->r_off;
+	uint64_t len = new->r_len;
+
+	/*
+	 * prev arrives either:
+	 * - pointing to an entry at the same offset
+	 * - pointing to the entry with the closest previous offset whose
+	 *   range may overlap with the new range
+	 * - null, if there were no ranges starting before the new one
+	 */
+	if (prev) {
+		if (prev->r_off + prev->r_len <= off) {
+			prev = NULL;
+		} else if (prev->r_off != off) {
+			/*
+			 * convert to proxy if needed then
+			 * split this entry and bump ref count
+			 */
+			prev = zfs_range_split(tree, prev, off);
+			prev = AVL_NEXT(tree, prev); /* move to rear range */
+		}
+	}
+	ASSERT((prev == NULL) || (prev->r_off == off));
+
+	if (prev)
+		next = prev;
+	else
+		next = (rl_t *)avl_nearest(tree, where, AVL_AFTER);
+
+	if (next == NULL || off + len <= next->r_off) {
+		/* no overlaps, use the original new rl_t in the tree */
+		avl_insert(tree, new, where);
+		return;
+	}
+
+	if (off < next->r_off) {
+		/* Add a proxy for initial range before the overlap */
+		zfs_range_new_proxy(tree, off, next->r_off - off);
+	}
+
+	new->r_cnt = 0; /* will use proxies in tree */
+	/*
+	 * We now search forward through the ranges, until we go past the end
+	 * of the new range. For each entry we make it a proxy if it
+	 * isn't already, then bump its reference count. If there's any
+	 * gaps between the ranges then we create a new proxy range.
+	 */
+	for (prev = NULL; next; prev = next, next = AVL_NEXT(tree, next)) {
+		if (off + len <= next->r_off)
+			break;
+		if (prev && prev->r_off + prev->r_len < next->r_off) {
+			/* there's a gap */
+			ASSERT3U(next->r_off, >, prev->r_off + prev->r_len);
+			zfs_range_new_proxy(tree, prev->r_off + prev->r_len,
+			    next->r_off - (prev->r_off + prev->r_len));
+		}
+		if (off + len == next->r_off + next->r_len) {
+			/* exact overlap with end */
+			next = zfs_range_proxify(tree, next);
+			next->r_cnt++;
+			return;
+		}
+		if (off + len < next->r_off + next->r_len) {
+			/* new range ends in the middle of this block */
+			next = zfs_range_split(tree, next, off + len);
+			next->r_cnt++;
+			return;
+		}
+		ASSERT3U(off + len, >, next->r_off + next->r_len);
+		next = zfs_range_proxify(tree, next);
+		next->r_cnt++;
+	}
+
+	/* Add the remaining end range. */
+	zfs_range_new_proxy(tree, prev->r_off + prev->r_len,
+	    (off + len) - (prev->r_off + prev->r_len));
+}
+
+/*
+ * Check if a reader lock can be grabbed, or wait and recheck until available.
+ */
+static void
+zfs_range_lock_reader(znode_t *zp, rl_t *new)
+{
+	avl_tree_t *tree = &zp->z_range_avl;
+	rl_t *prev, *next;
+	avl_index_t where;
+	uint64_t off = new->r_off;
+	uint64_t len = new->r_len;
+
+	/*
+	 * Look for any writer locks in the range.
+	 */
+retry:
+	prev = avl_find(tree, new, &where);
+	if (prev == NULL)
+		prev = (rl_t *)avl_nearest(tree, where, AVL_BEFORE);
+
+	/*
+	 * Check the previous range for a writer lock overlap.
+	 */
+	if (prev && (off < prev->r_off + prev->r_len)) {
+		if ((prev->r_type == RL_WRITER) || (prev->r_write_wanted)) {
+			if (!prev->r_read_wanted) {
+				cv_init(&prev->r_rd_cv, NULL, CV_DEFAULT, NULL);
+				prev->r_read_wanted = B_TRUE;
+			}
+			cv_wait(&prev->r_rd_cv, &zp->z_range_lock);
+			goto retry;
+		}
+		if (off + len < prev->r_off + prev->r_len)
+			goto got_lock;
+	}
+
+	/*
+	 * Search through the following ranges to see if there's
+	 * write lock any overlap.
+	 */
+	if (prev)
+		next = AVL_NEXT(tree, prev);
+	else
+		next = (rl_t *)avl_nearest(tree, where, AVL_AFTER);
+	for (; next; next = AVL_NEXT(tree, next)) {
+		if (off + len <= next->r_off)
+			goto got_lock;
+		if ((next->r_type == RL_WRITER) || (next->r_write_wanted)) {
+			if (!next->r_read_wanted) {
+				cv_init(&next->r_rd_cv, NULL, CV_DEFAULT, NULL);
+				next->r_read_wanted = B_TRUE;
+			}
+			cv_wait(&next->r_rd_cv, &zp->z_range_lock);
+			goto retry;
+		}
+		if (off + len <= next->r_off + next->r_len)
+			goto got_lock;
+	}
+
+got_lock:
+	/*
+	 * Add the read lock, which may involve splitting existing
+	 * locks and bumping ref counts (r_cnt).
+	 */
+	zfs_range_add_reader(tree, new, prev, where);
+}
+
+/*
+ * Lock a range (offset, length) as either shared (RL_READER)
+ * or exclusive (RL_WRITER). Returns the range lock structure
+ * for later unlocking or reduce range (if entire file
+ * previously locked as RL_WRITER).
+ */
+rl_t *
+zfs_range_lock(znode_t *zp, uint64_t off, uint64_t len, rl_type_t type)
+{
+	rl_t *new;
+
+	ASSERT(type == RL_READER || type == RL_WRITER || type == RL_APPEND);
+
+	new = kmem_alloc(sizeof (rl_t), KM_SLEEP);
+	new->r_zp = zp;
+	new->r_off = off;
+	new->r_len = len;
+	new->r_cnt = 1; /* assume it's going to be in the tree */
+	new->r_type = type;
+	new->r_proxy = B_FALSE;
+	new->r_write_wanted = B_FALSE;
+	new->r_read_wanted = B_FALSE;
+
+	mutex_enter(&zp->z_range_lock);
+	if (type == RL_READER) {
+		/*
+		 * First check for the usual case of no locks
+		 */
+		if (avl_numnodes(&zp->z_range_avl) == 0)
+			avl_add(&zp->z_range_avl, new);
+		else
+			zfs_range_lock_reader(zp, new);
+	} else
+		zfs_range_lock_writer(zp, new); /* RL_WRITER or RL_APPEND */
+	mutex_exit(&zp->z_range_lock);
+	return (new);
+}
+
+/*
+ * Unlock a reader lock
+ */
+static void
+zfs_range_unlock_reader(znode_t *zp, rl_t *remove)
+{
+	avl_tree_t *tree = &zp->z_range_avl;
+	rl_t *rl, *next;
+	uint64_t len;
+
+	/*
+	 * The common case is when the remove entry is in the tree
+	 * (cnt == 1) meaning there's been no other reader locks overlapping
+	 * with this one. Otherwise the remove entry will have been
+	 * removed from the tree and replaced by proxies (one or
+	 * more ranges mapping to the entire range).
+	 */
+	if (remove->r_cnt == 1) {
+		avl_remove(tree, remove);
+		if (remove->r_write_wanted) {
+			cv_broadcast(&remove->r_wr_cv);
+			cv_destroy(&remove->r_wr_cv);
+		}
+		if (remove->r_read_wanted) {
+			cv_broadcast(&remove->r_rd_cv);
+			cv_destroy(&remove->r_rd_cv);
+		}
+	} else {
+		ASSERT3U(remove->r_cnt, ==, 0);
+		ASSERT3U(remove->r_write_wanted, ==, 0);
+		ASSERT3U(remove->r_read_wanted, ==, 0);
+		/*
+		 * Find start proxy representing this reader lock,
+		 * then decrement ref count on all proxies
+		 * that make up this range, freeing them as needed.
+		 */
+		rl = avl_find(tree, remove, NULL);
+		ASSERT(rl);
+		ASSERT(rl->r_cnt);
+		ASSERT(rl->r_type == RL_READER);
+		for (len = remove->r_len; len != 0; rl = next) {
+			len -= rl->r_len;
+			if (len) {
+				next = AVL_NEXT(tree, rl);
+				ASSERT(next);
+				ASSERT(rl->r_off + rl->r_len == next->r_off);
+				ASSERT(next->r_cnt);
+				ASSERT(next->r_type == RL_READER);
+			}
+			rl->r_cnt--;
+			if (rl->r_cnt == 0) {
+				avl_remove(tree, rl);
+				if (rl->r_write_wanted) {
+					cv_broadcast(&rl->r_wr_cv);
+					cv_destroy(&rl->r_wr_cv);
+				}
+				if (rl->r_read_wanted) {
+					cv_broadcast(&rl->r_rd_cv);
+					cv_destroy(&rl->r_rd_cv);
+				}
+				kmem_free(rl, sizeof (rl_t));
+			}
+		}
+	}
+	kmem_free(remove, sizeof (rl_t));
+}
+
+/*
+ * Unlock range and destroy range lock structure.
+ */
+void
+zfs_range_unlock(rl_t *rl)
+{
+	znode_t *zp = rl->r_zp;
+
+	ASSERT(rl->r_type == RL_WRITER || rl->r_type == RL_READER);
+	ASSERT(rl->r_cnt == 1 || rl->r_cnt == 0);
+	ASSERT(!rl->r_proxy);
+
+	mutex_enter(&zp->z_range_lock);
+	if (rl->r_type == RL_WRITER) {
+		/* writer locks can't be shared or split */
+		avl_remove(&zp->z_range_avl, rl);
+		mutex_exit(&zp->z_range_lock);
+		if (rl->r_write_wanted) {
+			cv_broadcast(&rl->r_wr_cv);
+			cv_destroy(&rl->r_wr_cv);
+		}
+		if (rl->r_read_wanted) {
+			cv_broadcast(&rl->r_rd_cv);
+			cv_destroy(&rl->r_rd_cv);
+		}
+		kmem_free(rl, sizeof (rl_t));
+	} else {
+		/*
+		 * lock may be shared, let zfs_range_unlock_reader()
+		 * release the lock and free the rl_t
+		 */
+		zfs_range_unlock_reader(zp, rl);
+		mutex_exit(&zp->z_range_lock);
+	}
+}
+
+/*
+ * Reduce range locked as RL_WRITER from whole file to specified range.
+ * Asserts the whole file is exclusivly locked and so there's only one
+ * entry in the tree.
+ */
+void
+zfs_range_reduce(rl_t *rl, uint64_t off, uint64_t len)
+{
+	znode_t *zp = rl->r_zp;
+
+	/* Ensure there are no other locks */
+	ASSERT(avl_numnodes(&zp->z_range_avl) == 1);
+	ASSERT(rl->r_off == 0);
+	ASSERT(rl->r_type == RL_WRITER);
+	ASSERT(!rl->r_proxy);
+	ASSERT3U(rl->r_len, ==, UINT64_MAX);
+	ASSERT3U(rl->r_cnt, ==, 1);
+
+	mutex_enter(&zp->z_range_lock);
+	rl->r_off = off;
+	rl->r_len = len;
+	mutex_exit(&zp->z_range_lock);
+	if (rl->r_write_wanted)
+		cv_broadcast(&rl->r_wr_cv);
+	if (rl->r_read_wanted)
+		cv_broadcast(&rl->r_rd_cv);
+}
+
+/*
+ * AVL comparison function used to order range locks
+ * Locks are ordered on the start offset of the range.
+ */
+int
+zfs_range_compare(const void *arg1, const void *arg2)
+{
+	const rl_t *rl1 = arg1;
+	const rl_t *rl2 = arg2;
+
+	if (rl1->r_off > rl2->r_off)
+		return (1);
+	if (rl1->r_off < rl2->r_off)
+		return (-1);
+	return (0);
+}
--- a/zfs/lib/libdmu-ctl/zfs_vfsops.c
+++ b/zfs/lib/libdmu-ctl/zfs_vfsops.c
--- a/zfs/lib/libdmu-ctl/zfs_vnops.c
+++ b/zfs/lib/libdmu-ctl/zfs_vnops.c
--- a/zfs/lib/libdmu-ctl/zvol.c
+++ b/zfs/lib/libdmu-ctl/zvol.c
--- a/zfs/lib/libnvpair/Makefile.in
+++ b/zfs/lib/libnvpair/Makefile.in
@ -0,0 +1,34 @@
+subdir-m += include
+DISTFILES = libnvpair.c nvpair.c nvpair_alloc_fixed.c nvpair_alloc_system.c
+
+MODULE  := znvpair
+LIBRARY := libnvpair
+
+# Compile as kernel module.  Needed symlinks created for all
+# k* objects created by top level configure script.
+
+EXTRA_CFLAGS  = @KERNELCPPFLAGS@
+EXTRA_CFLAGS += -I@LIBDIR@/libnvpair/include
+
+obj-m := ${MODULE}.o
+
+${MODULE}-objs += knvpair.o		# Interfaces name/value pairs
+${MODULE}-objs += nvpair_alloc_spl.o	# Generic alloc/free support
+
+# Compile as shared library.  There's an extra useless host program
+# here called 'zu' because it was the easiest way I could convince
+# the kernel build system to construct a user space shared library.
+
+HOSTCFLAGS += @HOSTCFLAGS@
+HOSTCFLAGS += -I@LIBDIR@/libsolcompat/include
+HOSTCFLAGS += -I@LIBDIR@/libport/include
+HOSTCFLAGS += -I@LIBDIR@/libnvpair/include
+
+hostprogs-y := zu
+always := $(hostprogs-y)
+
+zu-objs := zu.o ${LIBRARY}.so
+
+${LIBRARY}-objs += unvpair.o
+${LIBRARY}-objs += nvpair_alloc_system.o
+${LIBRARY}-objs += libnvpair.o
--- a/zfs/lib/libnvpair/include/Makefile.in
+++ b/zfs/lib/libnvpair/include/Makefile.in
@ -0,0 +1,2 @@
+subdir-m += sys
+DISTFILES = libnvpair.h
--- a/zfs/lib/libnvpair/include/libnvpair.h
+++ b/zfs/lib/libnvpair/include/libnvpair.h
@ -0,0 +1,46 @@
+/*
+ * CDDL HEADER START
+ *
+ * The contents of this file are subject to the terms of the
+ * Common Development and Distribution License, Version 1.0 only
+ * (the "License").  You may not use this file except in compliance
+ * with the License.
+ *
+ * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
+ * or http://www.opensolaris.org/os/licensing.
+ * See the License for the specific language governing permissions
+ * and limitations under the License.
+ *
+ * When distributing Covered Code, include this CDDL HEADER in each
+ * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
+ * If applicable, add the following below this CDDL HEADER, with the
+ * fields enclosed by brackets "[]" replaced with your own identifying
+ * information: Portions Copyright [yyyy] [name of copyright owner]
+ *
+ * CDDL HEADER END
+ */
+/*
+ * Copyright 2005 Sun Microsystems, Inc.  All rights reserved.
+ * Use is subject to license terms.
+ */
+
+#ifndef	_LIBNVPAIR_H
+#define	_LIBNVPAIR_H
+
+
+
+#include <sys/nvpair.h>
+#include <stdlib.h>
+#include <stdio.h>
+
+#ifdef	__cplusplus
+extern "C" {
+#endif
+
+void nvlist_print(FILE *, nvlist_t *);
+
+#ifdef	__cplusplus
+}
+#endif
+
+#endif	/* _LIBNVPAIR_H */
--- a/zfs/lib/libnvpair/include/sys/Makefile.in
+++ b/zfs/lib/libnvpair/include/sys/Makefile.in
@ -0,0 +1 @@
+DISTFILES = nvpair.h nvpair_impl.h
--- a/zfs/lib/libnvpair/include/sys/nvpair.h
+++ b/zfs/lib/libnvpair/include/sys/nvpair.h
@ -0,0 +1,262 @@
+/*
+ * CDDL HEADER START
+ *
+ * The contents of this file are subject to the terms of the
+ * Common Development and Distribution License (the "License").
+ * You may not use this file except in compliance with the License.
+ *
+ * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
+ * or http://www.opensolaris.org/os/licensing.
+ * See the License for the specific language governing permissions
+ * and limitations under the License.
+ *
+ * When distributing Covered Code, include this CDDL HEADER in each
+ * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
+ * If applicable, add the following below this CDDL HEADER, with the
+ * fields enclosed by brackets "[]" replaced with your own identifying
+ * information: Portions Copyright [yyyy] [name of copyright owner]
+ *
+ * CDDL HEADER END
+ */
+/*
+ * Copyright 2007 Sun Microsystems, Inc.  All rights reserved.
+ * Use is subject to license terms.
+ */
+
+#ifndef	_SYS_NVPAIR_H
+#define	_SYS_NVPAIR_H
+
+
+
+#include <sys/types.h>
+#include <sys/errno.h>
+#include <sys/va_list.h>
+
+#if defined(_KERNEL) && !defined(_BOOT)
+#include <sys/kmem.h>
+#endif
+
+#ifdef	__cplusplus
+extern "C" {
+#endif
+
+typedef enum {
+	DATA_TYPE_UNKNOWN = 0,
+	DATA_TYPE_BOOLEAN,
+	DATA_TYPE_BYTE,
+	DATA_TYPE_INT16,
+	DATA_TYPE_UINT16,
+	DATA_TYPE_INT32,
+	DATA_TYPE_UINT32,
+	DATA_TYPE_INT64,
+	DATA_TYPE_UINT64,
+	DATA_TYPE_STRING,
+	DATA_TYPE_BYTE_ARRAY,
+	DATA_TYPE_INT16_ARRAY,
+	DATA_TYPE_UINT16_ARRAY,
+	DATA_TYPE_INT32_ARRAY,
+	DATA_TYPE_UINT32_ARRAY,
+	DATA_TYPE_INT64_ARRAY,
+	DATA_TYPE_UINT64_ARRAY,
+	DATA_TYPE_STRING_ARRAY,
+	DATA_TYPE_HRTIME,
+	DATA_TYPE_NVLIST,
+	DATA_TYPE_NVLIST_ARRAY,
+	DATA_TYPE_BOOLEAN_VALUE,
+	DATA_TYPE_INT8,
+	DATA_TYPE_UINT8,
+	DATA_TYPE_BOOLEAN_ARRAY,
+	DATA_TYPE_INT8_ARRAY,
+	DATA_TYPE_UINT8_ARRAY
+} data_type_t;
+
+typedef struct nvpair {
+	int32_t nvp_size;	/* size of this nvpair */
+	int16_t	nvp_name_sz;	/* length of name string */
+	int16_t	nvp_reserve;	/* not used */
+	int32_t	nvp_value_elem;	/* number of elements for array types */
+	data_type_t nvp_type;	/* type of value */
+	/* name string */
+	/* aligned ptr array for string arrays */
+	/* aligned array of data for value */
+} nvpair_t;
+
+/* nvlist header */
+typedef struct nvlist {
+	int32_t		nvl_version;
+	uint32_t	nvl_nvflag;	/* persistent flags */
+	uint64_t	nvl_priv;	/* ptr to private data if not packed */
+	uint32_t	nvl_flag;
+	int32_t		nvl_pad;	/* currently not used, for alignment */
+} nvlist_t;
+
+/* nvp implementation version */
+#define	NV_VERSION	0
+
+/* nvlist pack encoding */
+#define	NV_ENCODE_NATIVE	0
+#define	NV_ENCODE_XDR		1
+
+/* nvlist persistent unique name flags, stored in nvl_nvflags */
+#define	NV_UNIQUE_NAME		0x1
+#define	NV_UNIQUE_NAME_TYPE	0x2
+
+/* nvlist lookup pairs related flags */
+#define	NV_FLAG_NOENTOK		0x1
+
+/* convenience macros */
+#define	NV_ALIGN(x)		(((ulong_t)(x) + 7ul) & ~7ul)
+#define	NV_ALIGN4(x)		(((x) + 3) & ~3)
+
+#define	NVP_SIZE(nvp)		((nvp)->nvp_size)
+#define	NVP_NAME(nvp)		((char *)(nvp) + sizeof (nvpair_t))
+#define	NVP_TYPE(nvp)		((nvp)->nvp_type)
+#define	NVP_NELEM(nvp)		((nvp)->nvp_value_elem)
+#define	NVP_VALUE(nvp)		((char *)(nvp) + NV_ALIGN(sizeof (nvpair_t) \
+				+ (nvp)->nvp_name_sz))
+
+#define	NVL_VERSION(nvl)	((nvl)->nvl_version)
+#define	NVL_SIZE(nvl)		((nvl)->nvl_size)
+#define	NVL_FLAG(nvl)		((nvl)->nvl_flag)
+
+/* NV allocator framework */
+typedef struct nv_alloc_ops nv_alloc_ops_t;
+
+typedef struct nv_alloc {
+	const nv_alloc_ops_t *nva_ops;
+	void *nva_arg;
+} nv_alloc_t;
+
+struct nv_alloc_ops {
+	int (*nv_ao_init)(nv_alloc_t *, __va_list);
+	void (*nv_ao_fini)(nv_alloc_t *);
+	void *(*nv_ao_alloc)(nv_alloc_t *, size_t);
+	void (*nv_ao_free)(nv_alloc_t *, void *, size_t);
+	void (*nv_ao_reset)(nv_alloc_t *);
+};
+
+extern const nv_alloc_ops_t *nv_fixed_ops;
+extern nv_alloc_t *nv_alloc_nosleep;
+
+#if defined(_KERNEL) && !defined(_BOOT)
+extern nv_alloc_t *nv_alloc_sleep;
+#endif
+
+int nv_alloc_init(nv_alloc_t *, const nv_alloc_ops_t *, /* args */ ...);
+void nv_alloc_reset(nv_alloc_t *);
+void nv_alloc_fini(nv_alloc_t *);
+
+/* list management */
+int nvlist_alloc(nvlist_t **, uint_t, int);
+void nvlist_free(nvlist_t *);
+int nvlist_size(nvlist_t *, size_t *, int);
+int nvlist_pack(nvlist_t *, char **, size_t *, int, int);
+int nvlist_unpack(char *, size_t, nvlist_t **, int);
+int nvlist_dup(nvlist_t *, nvlist_t **, int);
+int nvlist_merge(nvlist_t *, nvlist_t *, int);
+
+int nvlist_xalloc(nvlist_t **, uint_t, nv_alloc_t *);
+int nvlist_xpack(nvlist_t *, char **, size_t *, int, nv_alloc_t *);
+int nvlist_xunpack(char *, size_t, nvlist_t **, nv_alloc_t *);
+int nvlist_xdup(nvlist_t *, nvlist_t **, nv_alloc_t *);
+nv_alloc_t *nvlist_lookup_nv_alloc(nvlist_t *);
+
+int nvlist_add_nvpair(nvlist_t *, nvpair_t *);
+int nvlist_add_boolean(nvlist_t *, const char *);
+int nvlist_add_boolean_value(nvlist_t *, const char *, boolean_t);
+int nvlist_add_byte(nvlist_t *, const char *, uchar_t);
+int nvlist_add_int8(nvlist_t *, const char *, int8_t);
+int nvlist_add_uint8(nvlist_t *, const char *, uint8_t);
+int nvlist_add_int16(nvlist_t *, const char *, int16_t);
+int nvlist_add_uint16(nvlist_t *, const char *, uint16_t);
+int nvlist_add_int32(nvlist_t *, const char *, int32_t);
+int nvlist_add_uint32(nvlist_t *, const char *, uint32_t);
+int nvlist_add_int64(nvlist_t *, const char *, int64_t);
+int nvlist_add_uint64(nvlist_t *, const char *, uint64_t);
+int nvlist_add_string(nvlist_t *, const char *, const char *);
+int nvlist_add_nvlist(nvlist_t *, const char *, nvlist_t *);
+int nvlist_add_boolean_array(nvlist_t *, const char *, boolean_t *, uint_t);
+int nvlist_add_byte_array(nvlist_t *, const char *, uchar_t *, uint_t);
+int nvlist_add_int8_array(nvlist_t *, const char *, int8_t *, uint_t);
+int nvlist_add_uint8_array(nvlist_t *, const char *, uint8_t *, uint_t);
+int nvlist_add_int16_array(nvlist_t *, const char *, int16_t *, uint_t);
+int nvlist_add_uint16_array(nvlist_t *, const char *, uint16_t *, uint_t);
+int nvlist_add_int32_array(nvlist_t *, const char *, int32_t *, uint_t);
+int nvlist_add_uint32_array(nvlist_t *, const char *, uint32_t *, uint_t);
+int nvlist_add_int64_array(nvlist_t *, const char *, int64_t *, uint_t);
+int nvlist_add_uint64_array(nvlist_t *, const char *, uint64_t *, uint_t);
+int nvlist_add_string_array(nvlist_t *, const char *, char *const *, uint_t);
+int nvlist_add_nvlist_array(nvlist_t *, const char *, nvlist_t **, uint_t);
+int nvlist_add_hrtime(nvlist_t *, const char *, hrtime_t);
+
+int nvlist_remove(nvlist_t *, const char *, data_type_t);
+int nvlist_remove_all(nvlist_t *, const char *);
+
+int nvlist_lookup_boolean(nvlist_t *, const char *);
+int nvlist_lookup_boolean_value(nvlist_t *, const char *, boolean_t *);
+int nvlist_lookup_byte(nvlist_t *, const char *, uchar_t *);
+int nvlist_lookup_int8(nvlist_t *, const char *, int8_t *);
+int nvlist_lookup_uint8(nvlist_t *, const char *, uint8_t *);
+int nvlist_lookup_int16(nvlist_t *, const char *, int16_t *);
+int nvlist_lookup_uint16(nvlist_t *, const char *, uint16_t *);
+int nvlist_lookup_int32(nvlist_t *, const char *, int32_t *);
+int nvlist_lookup_uint32(nvlist_t *, const char *, uint32_t *);
+int nvlist_lookup_int64(nvlist_t *, const char *, int64_t *);
+int nvlist_lookup_uint64(nvlist_t *, const char *, uint64_t *);
+int nvlist_lookup_string(nvlist_t *, const char *, char **);
+int nvlist_lookup_nvlist(nvlist_t *, const char *, nvlist_t **);
+int nvlist_lookup_boolean_array(nvlist_t *, const char *,
+    boolean_t **, uint_t *);
+int nvlist_lookup_byte_array(nvlist_t *, const char *, uchar_t **, uint_t *);
+int nvlist_lookup_int8_array(nvlist_t *, const char *, int8_t **, uint_t *);
+int nvlist_lookup_uint8_array(nvlist_t *, const char *, uint8_t **, uint_t *);
+int nvlist_lookup_int16_array(nvlist_t *, const char *, int16_t **, uint_t *);
+int nvlist_lookup_uint16_array(nvlist_t *, const char *, uint16_t **, uint_t *);
+int nvlist_lookup_int32_array(nvlist_t *, const char *, int32_t **, uint_t *);
+int nvlist_lookup_uint32_array(nvlist_t *, const char *, uint32_t **, uint_t *);
+int nvlist_lookup_int64_array(nvlist_t *, const char *, int64_t **, uint_t *);
+int nvlist_lookup_uint64_array(nvlist_t *, const char *, uint64_t **, uint_t *);
+int nvlist_lookup_string_array(nvlist_t *, const char *, char ***, uint_t *);
+int nvlist_lookup_nvlist_array(nvlist_t *, const char *,
+    nvlist_t ***, uint_t *);
+int nvlist_lookup_hrtime(nvlist_t *, const char *, hrtime_t *);
+int nvlist_lookup_pairs(nvlist_t *nvl, int, ...);
+
+int nvlist_lookup_nvpair(nvlist_t *nvl, const char *, nvpair_t **);
+boolean_t nvlist_exists(nvlist_t *nvl, const char *);
+
+/* processing nvpair */
+nvpair_t *nvlist_next_nvpair(nvlist_t *nvl, nvpair_t *);
+char *nvpair_name(nvpair_t *);
+data_type_t nvpair_type(nvpair_t *);
+int nvpair_value_boolean_value(nvpair_t *, boolean_t *);
+int nvpair_value_byte(nvpair_t *, uchar_t *);
+int nvpair_value_int8(nvpair_t *, int8_t *);
+int nvpair_value_uint8(nvpair_t *, uint8_t *);
+int nvpair_value_int16(nvpair_t *, int16_t *);
+int nvpair_value_uint16(nvpair_t *, uint16_t *);
+int nvpair_value_int32(nvpair_t *, int32_t *);
+int nvpair_value_uint32(nvpair_t *, uint32_t *);
+int nvpair_value_int64(nvpair_t *, int64_t *);
+int nvpair_value_uint64(nvpair_t *, uint64_t *);
+int nvpair_value_string(nvpair_t *, char **);
+int nvpair_value_nvlist(nvpair_t *, nvlist_t **);
+int nvpair_value_boolean_array(nvpair_t *, boolean_t **, uint_t *);
+int nvpair_value_byte_array(nvpair_t *, uchar_t **, uint_t *);
+int nvpair_value_int8_array(nvpair_t *, int8_t **, uint_t *);
+int nvpair_value_uint8_array(nvpair_t *, uint8_t **, uint_t *);
+int nvpair_value_int16_array(nvpair_t *, int16_t **, uint_t *);
+int nvpair_value_uint16_array(nvpair_t *, uint16_t **, uint_t *);
+int nvpair_value_int32_array(nvpair_t *, int32_t **, uint_t *);
+int nvpair_value_uint32_array(nvpair_t *, uint32_t **, uint_t *);
+int nvpair_value_int64_array(nvpair_t *, int64_t **, uint_t *);
+int nvpair_value_uint64_array(nvpair_t *, uint64_t **, uint_t *);
+int nvpair_value_string_array(nvpair_t *, char ***, uint_t *);
+int nvpair_value_nvlist_array(nvpair_t *, nvlist_t ***, uint_t *);
+int nvpair_value_hrtime(nvpair_t *, hrtime_t *);
+
+#ifdef	__cplusplus
+}
+#endif
+
+#endif	/* _SYS_NVPAIR_H */
--- a/zfs/lib/libnvpair/include/sys/nvpair_impl.h
+++ b/zfs/lib/libnvpair/include/sys/nvpair_impl.h
@ -0,0 +1,73 @@
+/*
+ * CDDL HEADER START
+ *
+ * The contents of this file are subject to the terms of the
+ * Common Development and Distribution License, Version 1.0 only
+ * (the "License").  You may not use this file except in compliance
+ * with the License.
+ *
+ * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
+ * or http://www.opensolaris.org/os/licensing.
+ * See the License for the specific language governing permissions
+ * and limitations under the License.
+ *
+ * When distributing Covered Code, include this CDDL HEADER in each
+ * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
+ * If applicable, add the following below this CDDL HEADER, with the
+ * fields enclosed by brackets "[]" replaced with your own identifying
+ * information: Portions Copyright [yyyy] [name of copyright owner]
+ *
+ * CDDL HEADER END
+ */
+/*
+ * Copyright 2004 Sun Microsystems, Inc.  All rights reserved.
+ * Use is subject to license terms.
+ */
+
+#ifndef	_NVPAIR_IMPL_H
+#define	_NVPAIR_IMPL_H
+
+
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+#include <sys/nvpair.h>
+
+/*
+ * The structures here provided for information and debugging purposes only
+ * may be changed in the future.
+ */
+
+/*
+ * implementation linked list for pre-packed data
+ */
+typedef struct i_nvp i_nvp_t;
+
+struct i_nvp {
+	union {
+		uint64_t	_nvi_align;	/* ensure alignment */
+		struct {
+			i_nvp_t	*_nvi_next;	/* pointer to next nvpair */
+			i_nvp_t	*_nvi_prev;	/* pointer to prev nvpair */
+		} _nvi;
+	} _nvi_un;
+	nvpair_t nvi_nvp;			/* nvpair */
+};
+#define	nvi_next	_nvi_un._nvi._nvi_next
+#define	nvi_prev	_nvi_un._nvi._nvi_prev
+
+typedef struct {
+	i_nvp_t		*nvp_list;	/* linked list of nvpairs */
+	i_nvp_t		*nvp_last;	/* last nvpair */
+	i_nvp_t		*nvp_curr;	/* current walker nvpair */
+	nv_alloc_t	*nvp_nva;	/* pluggable allocator */
+	uint32_t	nvp_stat;	/* internal state */
+} nvpriv_t;
+
+#ifdef	__cplusplus
+}
+#endif
+
+#endif	/* _NVPAIR_IMPL_H */
--- a/zfs/lib/libnvpair/libnvpair.c
+++ b/zfs/lib/libnvpair/libnvpair.c
@ -0,0 +1,266 @@
+/*
+ * CDDL HEADER START
+ *
+ * The contents of this file are subject to the terms of the
+ * Common Development and Distribution License, Version 1.0 only
+ * (the "License").  You may not use this file except in compliance
+ * with the License.
+ *
+ * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
+ * or http://www.opensolaris.org/os/licensing.
+ * See the License for the specific language governing permissions
+ * and limitations under the License.
+ *
+ * When distributing Covered Code, include this CDDL HEADER in each
+ * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
+ * If applicable, add the following below this CDDL HEADER, with the
+ * fields enclosed by brackets "[]" replaced with your own identifying
+ * information: Portions Copyright [yyyy] [name of copyright owner]
+ *
+ * CDDL HEADER END
+ */
+/*
+ * Copyright 2004 Sun Microsystems, Inc.  All rights reserved.
+ * Use is subject to license terms.
+ */
+
+
+
+#include <unistd.h>
+#include <strings.h>
+#include "libnvpair.h"
+
+/*
+ * libnvpair - A tools library for manipulating <name, value> pairs.
+ *
+ *	This library provides routines packing an unpacking nv pairs
+ *	for transporting data across process boundaries, transporting
+ *	between kernel and userland, and possibly saving onto disk files.
+ */
+
+static void
+indent(FILE *fp, int depth)
+{
+	while (depth-- > 0)
+		(void) fprintf(fp, "\t");
+}
+
+/*
+ * nvlist_print - Prints elements in an event buffer
+ */
+static
+void
+nvlist_print_with_indent(FILE *fp, nvlist_t *nvl, int depth)
+{
+	int i;
+	char *name;
+	uint_t nelem;
+	nvpair_t *nvp;
+
+	if (nvl == NULL)
+		return;
+
+	indent(fp, depth);
+	(void) fprintf(fp, "nvlist version: %d\n", NVL_VERSION(nvl));
+
+	nvp = nvlist_next_nvpair(nvl, NULL);
+
+	while (nvp) {
+		data_type_t type = nvpair_type(nvp);
+
+		indent(fp, depth);
+		name = nvpair_name(nvp);
+		(void) fprintf(fp, "\t%s =", name);
+		nelem = 0;
+		switch (type) {
+		case DATA_TYPE_BOOLEAN: {
+			(void) fprintf(fp, " 1");
+			break;
+		}
+		case DATA_TYPE_BOOLEAN_VALUE: {
+			boolean_t val;
+			(void) nvpair_value_boolean_value(nvp, &val);
+			(void) fprintf(fp, " %d", val);
+			break;
+		}
+		case DATA_TYPE_BYTE: {
+			uchar_t val;
+			(void) nvpair_value_byte(nvp, &val);
+			(void) fprintf(fp, " 0x%2.2x", val);
+			break;
+		}
+		case DATA_TYPE_INT8: {
+			int8_t val;
+			(void) nvpair_value_int8(nvp, &val);
+			(void) fprintf(fp, " %d", val);
+			break;
+		}
+		case DATA_TYPE_UINT8: {
+			uint8_t val;
+			(void) nvpair_value_uint8(nvp, &val);
+			(void) fprintf(fp, " 0x%x", val);
+			break;
+		}
+		case DATA_TYPE_INT16: {
+			int16_t val;
+			(void) nvpair_value_int16(nvp, &val);
+			(void) fprintf(fp, " %d", val);
+			break;
+		}
+		case DATA_TYPE_UINT16: {
+			uint16_t val;
+			(void) nvpair_value_uint16(nvp, &val);
+			(void) fprintf(fp, " 0x%x", val);
+			break;
+		}
+		case DATA_TYPE_INT32: {
+			int32_t val;
+			(void) nvpair_value_int32(nvp, &val);
+			(void) fprintf(fp, " %d", val);
+			break;
+		}
+		case DATA_TYPE_UINT32: {
+			uint32_t val;
+			(void) nvpair_value_uint32(nvp, &val);
+			(void) fprintf(fp, " 0x%x", val);
+			break;
+		}
+		case DATA_TYPE_INT64: {
+			int64_t val;
+			(void) nvpair_value_int64(nvp, &val);
+			(void) fprintf(fp, " %lld", (longlong_t)val);
+			break;
+		}
+		case DATA_TYPE_UINT64: {
+			uint64_t val;
+			(void) nvpair_value_uint64(nvp, &val);
+			(void) fprintf(fp, " 0x%llx", (u_longlong_t)val);
+			break;
+		}
+		case DATA_TYPE_STRING: {
+			char *val;
+			(void) nvpair_value_string(nvp, &val);
+			(void) fprintf(fp, " %s", val);
+			break;
+		}
+		case DATA_TYPE_BOOLEAN_ARRAY: {
+			boolean_t *val;
+			(void) nvpair_value_boolean_array(nvp, &val, &nelem);
+			for (i = 0; i < nelem; i++)
+				(void) fprintf(fp, " %d", val[i]);
+			break;
+		}
+		case DATA_TYPE_BYTE_ARRAY: {
+			uchar_t *val;
+			(void) nvpair_value_byte_array(nvp, &val, &nelem);
+			for (i = 0; i < nelem; i++)
+				(void) fprintf(fp, " 0x%2.2x", val[i]);
+			break;
+		}
+		case DATA_TYPE_INT8_ARRAY: {
+			int8_t *val;
+			(void) nvpair_value_int8_array(nvp, &val, &nelem);
+			for (i = 0; i < nelem; i++)
+				(void) fprintf(fp, " %d", val[i]);
+			break;
+		}
+		case DATA_TYPE_UINT8_ARRAY: {
+			uint8_t *val;
+			(void) nvpair_value_uint8_array(nvp, &val, &nelem);
+			for (i = 0; i < nelem; i++)
+				(void) fprintf(fp, " 0x%x", val[i]);
+			break;
+		}
+		case DATA_TYPE_INT16_ARRAY: {
+			int16_t *val;
+			(void) nvpair_value_int16_array(nvp, &val, &nelem);
+			for (i = 0; i < nelem; i++)
+				(void) fprintf(fp, " %d", val[i]);
+			break;
+		}
+		case DATA_TYPE_UINT16_ARRAY: {
+			uint16_t *val;
+			(void) nvpair_value_uint16_array(nvp, &val, &nelem);
+			for (i = 0; i < nelem; i++)
+				(void) fprintf(fp, " 0x%x", val[i]);
+			break;
+		}
+		case DATA_TYPE_INT32_ARRAY: {
+			int32_t *val;
+			(void) nvpair_value_int32_array(nvp, &val, &nelem);
+			for (i = 0; i < nelem; i++)
+				(void) fprintf(fp, " %d", val[i]);
+			break;
+		}
+		case DATA_TYPE_UINT32_ARRAY: {
+			uint32_t *val;
+			(void) nvpair_value_uint32_array(nvp, &val, &nelem);
+			for (i = 0; i < nelem; i++)
+				(void) fprintf(fp, " 0x%x", val[i]);
+			break;
+		}
+		case DATA_TYPE_INT64_ARRAY: {
+			int64_t *val;
+			(void) nvpair_value_int64_array(nvp, &val, &nelem);
+			for (i = 0; i < nelem; i++)
+				(void) fprintf(fp, " %lld", (longlong_t)val[i]);
+			break;
+		}
+		case DATA_TYPE_UINT64_ARRAY: {
+			uint64_t *val;
+			(void) nvpair_value_uint64_array(nvp, &val, &nelem);
+			for (i = 0; i < nelem; i++)
+				(void) fprintf(fp, " 0x%llx",
+				    (u_longlong_t)val[i]);
+			break;
+		}
+		case DATA_TYPE_STRING_ARRAY: {
+			char **val;
+			(void) nvpair_value_string_array(nvp, &val, &nelem);
+			for (i = 0; i < nelem; i++)
+				(void) fprintf(fp, " %s", val[i]);
+			break;
+		}
+		case DATA_TYPE_HRTIME: {
+			hrtime_t val;
+			(void) nvpair_value_hrtime(nvp, &val);
+			(void) fprintf(fp, " 0x%llx", val);
+			break;
+		}
+		case DATA_TYPE_NVLIST: {
+			nvlist_t *val;
+			(void) nvpair_value_nvlist(nvp, &val);
+			(void) fprintf(fp, " (embedded nvlist)\n");
+			nvlist_print_with_indent(fp, val, depth + 1);
+			indent(fp, depth + 1);
+			(void) fprintf(fp, "(end %s)\n", name);
+			break;
+		}
+		case DATA_TYPE_NVLIST_ARRAY: {
+			nvlist_t **val;
+			(void) nvpair_value_nvlist_array(nvp, &val, &nelem);
+			(void) fprintf(fp, " (array of embedded nvlists)\n");
+			for (i = 0; i < nelem; i++) {
+				indent(fp, depth + 1);
+				(void) fprintf(fp,
+				    "(start %s[%d])\n", name, i);
+				nvlist_print_with_indent(fp, val[i], depth + 1);
+				indent(fp, depth + 1);
+				(void) fprintf(fp, "(end %s[%d])\n", name, i);
+			}
+			break;
+		}
+		default:
+			(void) fprintf(fp, " unknown data type (%d)", type);
+			break;
+		}
+		(void) fprintf(fp, "\n");
+		nvp = nvlist_next_nvpair(nvl, nvp);
+	}
+}
+
+void
+nvlist_print(FILE *fp, nvlist_t *nvl)
+{
+	nvlist_print_with_indent(fp, nvl, 0);
+}
--- a/zfs/lib/libnvpair/nvpair.c
+++ b/zfs/lib/libnvpair/nvpair.c
--- a/zfs/lib/libnvpair/nvpair_alloc_fixed.c
+++ b/zfs/lib/libnvpair/nvpair_alloc_fixed.c
@ -0,0 +1,120 @@
+/*
+ * CDDL HEADER START
+ *
+ * The contents of this file are subject to the terms of the
+ * Common Development and Distribution License (the "License").
+ * You may not use this file except in compliance with the License.
+ *
+ * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
+ * or http://www.opensolaris.org/os/licensing.
+ * See the License for the specific language governing permissions
+ * and limitations under the License.
+ *
+ * When distributing Covered Code, include this CDDL HEADER in each
+ * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
+ * If applicable, add the following below this CDDL HEADER, with the
+ * fields enclosed by brackets "[]" replaced with your own identifying
+ * information: Portions Copyright [yyyy] [name of copyright owner]
+ *
+ * CDDL HEADER END
+ */
+
+/*
+ * Copyright 2006 Sun Microsystems, Inc.  All rights reserved.
+ * Use is subject to license terms.
+ */
+
+
+
+#include <sys/stropts.h>
+#include <sys/isa_defs.h>
+#include <sys/nvpair.h>
+#include <sys/sysmacros.h>
+#if defined(_KERNEL) && !defined(_BOOT)
+#include <sys/varargs.h>
+#else
+#include <stdarg.h>
+#include <strings.h>
+#endif
+
+/*
+ * This allocator is very simple.
+ *  - it uses a pre-allocated buffer for memory allocations.
+ *  - it does _not_ free memory in the pre-allocated buffer.
+ *
+ * The reason for the selected implemention is simplicity.
+ * This allocator is designed for the usage in interrupt context when
+ * the caller may not wait for free memory.
+ */
+
+/* pre-allocated buffer for memory allocations */
+typedef struct nvbuf {
+	uintptr_t	nvb_buf;	/* address of pre-allocated buffer */
+	uintptr_t 	nvb_lim;	/* limit address in the buffer */
+	uintptr_t	nvb_cur;	/* current address in the buffer */
+} nvbuf_t;
+
+/*
+ * Initialize the pre-allocated buffer allocator. The caller needs to supply
+ *
+ *   buf	address of pre-allocated buffer
+ *   bufsz	size of pre-allocated buffer
+ *
+ * nv_fixed_init() calculates the remaining members of nvbuf_t.
+ */
+static int
+nv_fixed_init(nv_alloc_t *nva, va_list valist)
+{
+	uintptr_t base = va_arg(valist, uintptr_t);
+	uintptr_t lim = base + va_arg(valist, size_t);
+	nvbuf_t *nvb = (nvbuf_t *)P2ROUNDUP(base, sizeof (uintptr_t));
+
+	if (base == 0 || (uintptr_t)&nvb[1] > lim)
+		return (EINVAL);
+
+	nvb->nvb_buf = (uintptr_t)&nvb[0];
+	nvb->nvb_cur = (uintptr_t)&nvb[1];
+	nvb->nvb_lim = lim;
+	nva->nva_arg = nvb;
+
+	return (0);
+}
+
+static void *
+nv_fixed_alloc(nv_alloc_t *nva, size_t size)
+{
+	nvbuf_t *nvb = nva->nva_arg;
+	uintptr_t new = nvb->nvb_cur;
+
+	if (size == 0 || new + size > nvb->nvb_lim)
+		return (NULL);
+
+	nvb->nvb_cur = P2ROUNDUP(new + size, sizeof (uintptr_t));
+
+	return ((void *)new);
+}
+
+/*ARGSUSED*/
+static void
+nv_fixed_free(nv_alloc_t *nva, void *buf, size_t size)
+{
+	/* don't free memory in the pre-allocated buffer */
+}
+
+static void
+nv_fixed_reset(nv_alloc_t *nva)
+{
+	nvbuf_t *nvb = nva->nva_arg;
+
+	nvb->nvb_cur = (uintptr_t)&nvb[1];
+}
+
+const nv_alloc_ops_t nv_fixed_ops_def = {
+	nv_fixed_init,	/* nv_ao_init() */
+	NULL,		/* nv_ao_fini() */
+	nv_fixed_alloc,	/* nv_ao_alloc() */
+	nv_fixed_free,	/* nv_ao_free() */
+	nv_fixed_reset	/* nv_ao_reset() */
+};
+
+const nv_alloc_ops_t *nv_fixed_ops = &nv_fixed_ops_def;
--- a/zfs/lib/libnvpair/nvpair_alloc_system.c
+++ b/zfs/lib/libnvpair/nvpair_alloc_system.c
@ -0,0 +1,59 @@
+/*
+ * CDDL HEADER START
+ *
+ * The contents of this file are subject to the terms of the
+ * Common Development and Distribution License, Version 1.0 only
+ * (the "License").  You may not use this file except in compliance
+ * with the License.
+ *
+ * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
+ * or http://www.opensolaris.org/os/licensing.
+ * See the License for the specific language governing permissions
+ * and limitations under the License.
+ *
+ * When distributing Covered Code, include this CDDL HEADER in each
+ * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
+ * If applicable, add the following below this CDDL HEADER, with the
+ * fields enclosed by brackets "[]" replaced with your own identifying
+ * information: Portions Copyright [yyyy] [name of copyright owner]
+ *
+ * CDDL HEADER END
+ */
+/*
+ * Copyright 2004 Sun Microsystems, Inc.  All rights reserved.
+ * Use is subject to license terms.
+ */
+
+
+
+#include <sys/nvpair.h>
+#include <stdlib.h>
+
+/*ARGSUSED*/
+static void *
+nv_alloc_sys(nv_alloc_t *nva, size_t size)
+{
+	return (malloc(size));
+}
+
+/*ARGSUSED*/
+static void
+nv_free_sys(nv_alloc_t *nva, void *buf, size_t size)
+{
+	free(buf);
+}
+
+const nv_alloc_ops_t system_ops_def = {
+	NULL,			/* nv_ao_init() */
+	NULL,			/* nv_ao_fini() */
+	nv_alloc_sys,		/* nv_ao_alloc() */
+	nv_free_sys,		/* nv_ao_free() */
+	NULL			/* nv_ao_reset() */
+};
+
+nv_alloc_t nv_alloc_nosleep_def = {
+	&system_ops_def,
+	NULL
+};
+
+nv_alloc_t *nv_alloc_nosleep = &nv_alloc_nosleep_def;
--- a/zfs/lib/libport/Makefile.in
+++ b/zfs/lib/libport/Makefile.in
@ -0,0 +1,37 @@
+subdir-m += include
+DISTFILES = port.c strlcat.c strlcpy.c strnlen.c u8_textprep.c
+
+MODULE  := zport
+LIBRARY := libzport
+
+# Compile as kernel module.  Needed symlinks created for all
+# k* objects created by top level configure script.
+
+EXTRA_CFLAGS  = @KERNELCPPFLAGS@
+EXTRA_CFLAGS += -I@LIBDIR@/libzcommon/include
+EXTRA_CFLAGS += -I@LIBDIR@/libport/include
+
+obj-m := ${MODULE}.o
+
+${MODULE}-objs += spl.o
+${MODULE}-objs += ku8_textprep.o
+
+# Compile as shared library.  There's an extra useless host program
+# here called 'zu' because it was the easiest way I could convince
+# the kernel build system to construct a user space shared library.
+
+HOSTCFLAGS += @HOSTCFLAGS@
+HOSTCFLAGS += -I@LIBDIR@/libzcommon/include
+HOSTCFLAGS += -I@LIBDIR@/libport/include
+
+hostprogs-y := zu
+always := $(hostprogs-y)
+
+zu-objs := zu.o ${LIBRARY}.so
+
+${LIBRARY}-objs += strlcpy.o
+${LIBRARY}-objs += strlcat.o
+${LIBRARY}-objs += strnlen.o
+${LIBRARY}-objs += port.o
+${LIBRARY}-objs += u8_textprep.o
+
--- a/zfs/lib/libport/include/Makefile.in
+++ b/zfs/lib/libport/include/Makefile.in
@ -0,0 +1,4 @@
+subdir-m += sys
+
+DISTFILES  = fake_ioctl.h libdiskmgt.h libshare.h mntent.h stdlib.h
+DISTFILES += string.h strings.h stropts.h unistd.h
--- a/zfs/lib/libport/include/fake_ioctl.h
+++ b/zfs/lib/libport/include/fake_ioctl.h
@ -0,0 +1,41 @@
+/*
+ * CDDL HEADER START
+ *
+ * The contents of this file are subject to the terms of the
+ * Common Development and Distribution License (the "License").
+ * You may not use this file except in compliance with the License.
+ *
+ * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
+ * or http://www.opensolaris.org/os/licensing.
+ * See the License for the specific language governing permissions
+ * and limitations under the License.
+ *
+ * When distributing Covered Code, include this CDDL HEADER in each
+ * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
+ * If applicable, add the following below this CDDL HEADER, with the
+ * fields enclosed by brackets "[]" replaced with your own identifying
+ * information: Portions Copyright [yyyy] [name of copyright owner]
+ *
+ * CDDL HEADER END
+ */
+/*
+ * Copyright 2008 Sun Microsystems, Inc.  All rights reserved.
+ * Use is subject to license terms.
+ */
+
+#ifndef _PORT_FAKE_IOCTL_H
+#define _PORT_FAKE_IOCTL_H
+
+static inline int real_ioctl(int fd, int request, void *arg)
+{
+	return ioctl(fd, request, arg);
+}
+
+#ifdef WANT_FAKE_IOCTL
+
+#include <sys/dmu_ctl.h>
+#define ioctl(fd,req,arg) dctlc_ioctl(fd,req,arg)
+
+#endif
+
+#endif /* _PORT_FAKE_IOCTL_H */
--- a/zfs/lib/libport/include/libdiskmgt.h
+++ b/zfs/lib/libport/include/libdiskmgt.h
@ -0,0 +1,278 @@
+/*
+ * CDDL HEADER START
+ *
+ * The contents of this file are subject to the terms of the
+ * Common Development and Distribution License (the "License").
+ * You may not use this file except in compliance with the License.
+ *
+ * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
+ * or http://www.opensolaris.org/os/licensing.
+ * See the License for the specific language governing permissions
+ * and limitations under the License.
+ *
+ * When distributing Covered Code, include this CDDL HEADER in each
+ * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
+ * If applicable, add the following below this CDDL HEADER, with the
+ * fields enclosed by brackets "[]" replaced with your own identifying
+ * information: Portions Copyright [yyyy] [name of copyright owner]
+ *
+ * CDDL HEADER END
+ */
+/*
+ * Copyright 2007 Sun Microsystems, Inc.  All rights reserved.
+ * Use is subject to license terms.
+ */
+
+#include "zfs_config.h"
+
+#ifdef HAVE_LIBDISKMGT_H
+#include_next <libdiskmgt.h>
+#else
+
+#ifndef _LIBDISKMGT_H
+#define	_LIBDISKMGT_H
+
+
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+#include <libnvpair.h>
+#include <sys/swap.h>
+
+
+/*
+ * Holds all the data regarding the device.
+ * Private to libdiskmgt. Must use dm_xxx functions to set/get data.
+ */
+typedef uint64_t  dm_descriptor_t;
+
+typedef enum {
+	DM_WHO_MKFS = 0,
+	DM_WHO_ZPOOL,
+	DM_WHO_ZPOOL_FORCE,
+	DM_WHO_FORMAT,
+	DM_WHO_SWAP,
+	DM_WHO_DUMP,
+	DM_WHO_ZPOOL_SPARE
+} dm_who_type_t;
+
+typedef enum {
+    DM_DRIVE = 0,
+    DM_CONTROLLER,
+    DM_MEDIA,
+    DM_SLICE,
+    DM_PARTITION,
+    DM_PATH,
+    DM_ALIAS,
+    DM_BUS
+} dm_desc_type_t;
+
+
+typedef enum {
+    DM_DT_UNKNOWN = 0,
+    DM_DT_FIXED,
+    DM_DT_ZIP,
+    DM_DT_JAZ,
+    DM_DT_FLOPPY,
+    DM_DT_MO_ERASABLE,
+    DM_DT_MO_WRITEONCE,
+    DM_DT_AS_MO,
+    DM_DT_CDROM,
+    DM_DT_CDR,
+    DM_DT_CDRW,
+    DM_DT_DVDROM,
+    DM_DT_DVDR,
+    DM_DT_DVDRAM,
+    DM_DT_DVDRW,
+    DM_DT_DDCDROM,
+    DM_DT_DDCDR,
+    DM_DT_DDCDRW
+} dm_drive_type_t;
+
+typedef enum {
+    DM_MT_UNKNOWN = 0,
+    DM_MT_FIXED,
+    DM_MT_FLOPPY,
+    DM_MT_CDROM,
+    DM_MT_ZIP,
+    DM_MT_JAZ,
+    DM_MT_CDR,
+    DM_MT_CDRW,
+    DM_MT_DVDROM,
+    DM_MT_DVDR,
+    DM_MT_DVDRAM,
+    DM_MT_MO_ERASABLE,
+    DM_MT_MO_WRITEONCE,
+    DM_MT_AS_MO
+} dm_media_type_t;
+
+#define	DM_FILTER_END	-1
+
+/* drive stat name */
+typedef enum {
+    DM_DRV_STAT_PERFORMANCE = 0,
+    DM_DRV_STAT_DIAGNOSTIC,
+    DM_DRV_STAT_TEMPERATURE
+} dm_drive_stat_t;
+
+/* slice stat name */
+typedef enum {
+    DM_SLICE_STAT_USE = 0
+} dm_slice_stat_t;
+
+/* attribute definitions */
+
+/* drive */
+#define	DM_DISK_UP		1
+#define	DM_DISK_DOWN		0
+
+#define	DM_CLUSTERED		"clustered"
+#define	DM_DRVTYPE		"drvtype"
+#define	DM_FAILING		"failing"
+#define	DM_LOADED		"loaded"	/* also in media */
+#define	DM_NDNRERRS		"ndevice_not_ready_errors"
+#define	DM_NBYTESREAD		"nbytes_read"
+#define	DM_NBYTESWRITTEN	"nbytes_written"
+#define	DM_NHARDERRS		"nhard_errors"
+#define	DM_NILLREQERRS		"nillegal_req_errors"
+#define	DM_NMEDIAERRS		"nmedia_errors"
+#define	DM_NNODEVERRS		"nno_dev_errors"
+#define	DM_NREADOPS		"nread_ops"
+#define	DM_NRECOVERRS		"nrecoverable_errors"
+#define	DM_NSOFTERRS		"nsoft_errors"
+#define	DM_NTRANSERRS		"ntransport_errors"
+#define	DM_NWRITEOPS		"nwrite_ops"
+#define	DM_OPATH		"opath"
+#define	DM_PRODUCT_ID		"product_id"
+#define	DM_REMOVABLE		"removable"	/* also in media */
+#define	DM_RPM			"rpm"
+#define	DM_STATUS		"status"
+#define	DM_SYNC_SPEED		"sync_speed"
+#define	DM_TEMPERATURE		"temperature"
+#define	DM_VENDOR_ID		"vendor_id"
+#define	DM_WIDE			"wide"		/* also on controller */
+#define	DM_WWN			"wwn"
+
+/* bus */
+#define	DM_BTYPE		"btype"
+#define	DM_CLOCK		"clock"		/* also on controller */
+#define	DM_PNAME		"pname"
+
+/* controller */
+#define	DM_FAST			"fast"
+#define	DM_FAST20		"fast20"
+#define	DM_FAST40		"fast40"
+#define	DM_FAST80		"fast80"
+#define	DM_MULTIPLEX		"multiplex"
+#define	DM_PATH_STATE		"path_state"
+
+#define	DM_CTYPE_ATA		"ata"
+#define	DM_CTYPE_SCSI		"scsi"
+#define	DM_CTYPE_FIBRE		"fibre channel"
+#define	DM_CTYPE_USB		"usb"
+#define	DM_CTYPE_UNKNOWN	"unknown"
+
+/* media */
+#define	DM_BLOCKSIZE		"blocksize"
+#define	DM_FDISK		"fdisk"
+#define	DM_MTYPE		"mtype"
+#define	DM_NACTUALCYLINDERS	"nactual_cylinders"
+#define	DM_NALTCYLINDERS	"nalt_cylinders"
+#define	DM_NCYLINDERS		"ncylinders"
+#define	DM_NHEADS		"nheads"
+#define	DM_NPHYSCYLINDERS	"nphys_cylinders"
+#define	DM_NSECTORS		"nsectors"	/* also in partition */
+#define	DM_SIZE			"size"		/* also in slice */
+#define	DM_NACCESSIBLE		"naccessible"
+#define	DM_LABEL		"label"
+
+/* partition */
+#define	DM_BCYL			"bcyl"
+#define	DM_BHEAD		"bhead"
+#define	DM_BOOTID		"bootid"
+#define	DM_BSECT		"bsect"
+#define	DM_ECYL			"ecyl"
+#define	DM_EHEAD		"ehead"
+#define	DM_ESECT		"esect"
+#define	DM_PTYPE		"ptype"
+#define	DM_RELSECT		"relsect"
+
+/* slice */
+#define	DM_DEVICEID		"deviceid"
+#define	DM_DEVT			"devt"
+#define	DM_INDEX		"index"
+#define	DM_EFI_NAME		"name"
+#define	DM_MOUNTPOINT		"mountpoint"
+#define	DM_LOCALNAME		"localname"
+#define	DM_START		"start"
+#define	DM_TAG			"tag"
+#define	DM_FLAG			"flag"
+#define	DM_EFI			"efi"	/* also on media */
+#define	DM_USED_BY		"used_by"
+#define	DM_USED_NAME		"used_name"
+#define	DM_USE_MOUNT		"mount"
+#define	DM_USE_SVM		"svm"
+#define	DM_USE_LU		"lu"
+#define	DM_USE_DUMP		"dump"
+#define	DM_USE_VXVM		"vxvm"
+#define	DM_USE_FS		"fs"
+#define	DM_USE_VFSTAB		"vfstab"
+#define	DM_USE_EXPORTED_ZPOOL	"exported_zpool"
+#define	DM_USE_ACTIVE_ZPOOL	"active_zpool"
+#define	DM_USE_SPARE_ZPOOL	"spare_zpool"
+#define	DM_USE_L2CACHE_ZPOOL	"l2cache_zpool"
+
+/* event */
+#define	DM_EV_NAME		"name"
+#define	DM_EV_DTYPE		"edtype"
+#define	DM_EV_TYPE		"evtype"
+#define	DM_EV_TADD		"add"
+#define	DM_EV_TREMOVE		"remove"
+#define	DM_EV_TCHANGE		"change"
+
+/* findisks */
+#define	DM_CTYPE		"ctype"
+#define	DM_LUN			"lun"
+#define	DM_TARGET		"target"
+
+#define	NOINUSE_SET	getenv("NOINUSE_CHECK") != NULL
+
+void			dm_free_descriptors(dm_descriptor_t *desc_list);
+void			dm_free_descriptor(dm_descriptor_t desc);
+void			dm_free_name(char *name);
+void			dm_free_swapentries(swaptbl_t *);
+
+dm_descriptor_t		*dm_get_descriptors(dm_desc_type_t type, int filter[],
+			    int *errp);
+dm_descriptor_t		*dm_get_associated_descriptors(dm_descriptor_t desc,
+			    dm_desc_type_t type, int *errp);
+dm_desc_type_t		*dm_get_associated_types(dm_desc_type_t type);
+dm_descriptor_t		dm_get_descriptor_by_name(dm_desc_type_t desc_type,
+			    char *name, int *errp);
+char			*dm_get_name(dm_descriptor_t desc, int *errp);
+dm_desc_type_t		dm_get_type(dm_descriptor_t desc);
+nvlist_t		*dm_get_attributes(dm_descriptor_t desc, int *errp);
+nvlist_t		*dm_get_stats(dm_descriptor_t desc, int stat_type,
+			    int *errp);
+void			dm_init_event_queue(void(*callback)(nvlist_t *, int),
+			    int *errp);
+nvlist_t		*dm_get_event(int *errp);
+void			dm_get_slices(char *drive, dm_descriptor_t **slices,
+			    int *errp);
+void			dm_get_slice_stats(char *slice, nvlist_t **dev_stats,
+			    int *errp);
+int			dm_get_swapentries(swaptbl_t **, int *);
+void			dm_get_usage_string(char *who, char *data, char **msg);
+int			dm_inuse(char *dev_name, char **msg, dm_who_type_t who,
+			    int *errp);
+int			dm_inuse_swap(const char *dev_name, int *errp);
+int			dm_isoverlapping(char *dev_name, char **msg, int *errp);
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif /* _LIBDISKMGT_H */
+#endif /* HAVE_LIBDISKMGT_H */
--- a/zfs/lib/libport/include/libshare.h
+++ b/zfs/lib/libport/include/libshare.h
@ -0,0 +1,287 @@
+/*
+ * CDDL HEADER START
+ *
+ * The contents of this file are subject to the terms of the
+ * Common Development and Distribution License (the "License").
+ * You may not use this file except in compliance with the License.
+ *
+ * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
+ * or http://www.opensolaris.org/os/licensing.
+ * See the License for the specific language governing permissions
+ * and limitations under the License.
+ *
+ * When distributing Covered Code, include this CDDL HEADER in each
+ * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
+ * If applicable, add the following below this CDDL HEADER, with the
+ * fields enclosed by brackets "[]" replaced with your own identifying
+ * information: Portions Copyright [yyyy] [name of copyright owner]
+ *
+ * CDDL HEADER END
+ */
+
+/*
+ * Copyright 2008 Sun Microsystems, Inc.  All rights reserved.
+ * Use is subject to license terms.
+ */
+
+/*
+ * basic API declarations for share management
+ */
+
+#include "zfs_config.h"
+
+#ifdef HAVE_LIBSHARE
+#include_next <libshare.h>
+#else
+
+#ifndef _LIBSHARE_H
+#define	_LIBSHARE_H
+
+
+
+#ifdef	__cplusplus
+extern "C" {
+#endif
+
+#include <sys/types.h>
+
+/*
+ * Basic datatypes for most functions
+ */
+typedef void *sa_group_t;
+typedef void *sa_share_t;
+typedef void *sa_property_t;
+typedef void *sa_optionset_t;
+typedef void *sa_security_t;
+typedef void *sa_protocol_properties_t;
+typedef void *sa_resource_t;
+
+typedef void *sa_handle_t;	/* opaque handle to access core functions */
+
+/*
+ * defined error values
+ */
+
+#define	SA_OK			0
+#define	SA_NO_SUCH_PATH		1	/* provided path doesn't exist */
+#define	SA_NO_MEMORY		2	/* no memory for data structures */
+#define	SA_DUPLICATE_NAME	3	/* object name is already in use */
+#define	SA_BAD_PATH		4	/* not a full path */
+#define	SA_NO_SUCH_GROUP	5	/* group is not defined */
+#define	SA_CONFIG_ERR		6	/* system configuration error */
+#define	SA_SYSTEM_ERR		7	/* system error, use errno */
+#define	SA_SYNTAX_ERR		8	/* syntax error on command line */
+#define	SA_NO_PERMISSION	9	/* no permission for operation */
+#define	SA_BUSY			10	/* resource is busy */
+#define	SA_NO_SUCH_PROP		11	/* property doesn't exist */
+#define	SA_INVALID_NAME		12	/* name of object is invalid */
+#define	SA_INVALID_PROTOCOL	13	/* specified protocol not valid */
+#define	SA_NOT_ALLOWED		14	/* operation not allowed */
+#define	SA_BAD_VALUE		15	/* bad value for property */
+#define	SA_INVALID_SECURITY	16	/* invalid security type */
+#define	SA_NO_SUCH_SECURITY	17	/* security set not found */
+#define	SA_VALUE_CONFLICT	18	/* property value conflict */
+#define	SA_NOT_IMPLEMENTED	19	/* plugin interface not implemented */
+#define	SA_INVALID_PATH		20	/* path is sub-dir of existing share */
+#define	SA_NOT_SUPPORTED	21	/* operation not supported for proto */
+#define	SA_PROP_SHARE_ONLY	22	/* property valid on share only */
+#define	SA_NOT_SHARED		23	/* path is not shared */
+#define	SA_NO_SUCH_RESOURCE	24	/* resource not found */
+#define	SA_RESOURCE_REQUIRED	25	/* resource name is required  */
+#define	SA_MULTIPLE_ERROR	26	/* multiple protocols reported error */
+#define	SA_PATH_IS_SUBDIR	27	/* check_path found path is subdir */
+#define	SA_PATH_IS_PARENTDIR	28	/* check_path found path is parent */
+#define	SA_NO_SECTION		29	/* protocol requires section info */
+#define	SA_NO_SUCH_SECTION	30	/* no section found */
+#define	SA_NO_PROPERTIES	31	/* no properties found */
+#define	SA_PASSWORD_ENC		32	/* passwords must be encrypted */
+
+/* API Initialization */
+#define	SA_INIT_SHARE_API	0x0001	/* init share specific interface */
+#define	SA_INIT_CONTROL_API	0x0002	/* init control specific interface */
+
+/* not part of API returns */
+#define	SA_LEGACY_ERR		32	/* share/unshare error return */
+
+/*
+ * other defined values
+ */
+
+#define	SA_MAX_NAME_LEN		100	/* must fit service instance name */
+#define	SA_MAX_RESOURCE_NAME	255	/* Maximum length of resource name */
+
+/* Used in calls to sa_add_share() and sa_add_resource() */
+#define	SA_SHARE_TRANSIENT	0	/* shared but not across reboot */
+#define	SA_SHARE_LEGACY		1	/* share is in dfstab only */
+#define	SA_SHARE_PERMANENT	2	/* share goes to repository */
+
+/* sa_check_path() related */
+#define	SA_CHECK_NORMAL		0	/* only check against active shares */
+#define	SA_CHECK_STRICT		1	/* check against all shares */
+
+/* RBAC related */
+#define	SA_RBAC_MANAGE	"solaris.smf.manage.shares"
+#define	SA_RBAC_VALUE	"solaris.smf.value.shares"
+
+/*
+ * Feature set bit definitions
+ */
+
+#define	SA_FEATURE_NONE		0x0000	/* no feature flags set */
+#define	SA_FEATURE_RESOURCE	0x0001	/* resource names are required */
+#define	SA_FEATURE_DFSTAB	0x0002	/* need to manage in dfstab */
+#define	SA_FEATURE_ALLOWSUBDIRS	0x0004	/* allow subdirs to be shared */
+#define	SA_FEATURE_ALLOWPARDIRS	0x0008	/* allow parent dirs to be shared */
+#define	SA_FEATURE_HAS_SECTIONS	0x0010	/* protocol supports sections */
+#define	SA_FEATURE_ADD_PROPERTIES	0x0020	/* can add properties */
+#define	SA_FEATURE_SERVER	0x0040	/* protocol supports server mode */
+
+/*
+ * legacy files
+ */
+
+#define	SA_LEGACY_DFSTAB	"/etc/dfs/dfstab"
+#define	SA_LEGACY_SHARETAB	"/etc/dfs/sharetab"
+
+/*
+ * SMF related
+ */
+
+#define	SA_SVC_FMRI_BASE	"svc:/network/shares/group"
+
+/* initialization */
+extern sa_handle_t sa_init(int);
+extern void sa_fini(sa_handle_t);
+extern int sa_update_config(sa_handle_t);
+extern char *sa_errorstr(int);
+
+/* protocol names */
+extern int sa_get_protocols(char ***);
+extern int sa_valid_protocol(char *);
+
+/* group control (create, remove, etc) */
+extern sa_group_t sa_create_group(sa_handle_t, char *, int *);
+extern int sa_remove_group(sa_group_t);
+extern sa_group_t sa_get_group(sa_handle_t, char *);
+extern sa_group_t sa_get_next_group(sa_group_t);
+extern char *sa_get_group_attr(sa_group_t, char *);
+extern int sa_set_group_attr(sa_group_t, char *, char *);
+extern sa_group_t sa_get_sub_group(sa_group_t);
+extern int sa_valid_group_name(char *);
+
+/* share control */
+extern sa_share_t sa_add_share(sa_group_t, char *, int, int *);
+extern int sa_check_path(sa_group_t, char *, int);
+extern int sa_move_share(sa_group_t, sa_share_t);
+extern int sa_remove_share(sa_share_t);
+extern sa_share_t sa_get_share(sa_group_t, char *);
+extern sa_share_t sa_find_share(sa_handle_t, char *);
+extern sa_share_t sa_get_next_share(sa_share_t);
+extern char *sa_get_share_attr(sa_share_t, char *);
+extern char *sa_get_share_description(sa_share_t);
+extern sa_group_t sa_get_parent_group(sa_share_t);
+extern int sa_set_share_attr(sa_share_t, char *, char *);
+extern int sa_set_share_description(sa_share_t, char *);
+extern int sa_enable_share(sa_group_t, char *);
+extern int sa_disable_share(sa_share_t, char *);
+extern int sa_is_share(void *);
+
+/* resource name related */
+extern sa_resource_t sa_find_resource(sa_handle_t, char *);
+extern sa_resource_t sa_get_resource(sa_group_t, char *);
+extern sa_resource_t sa_get_next_resource(sa_resource_t);
+extern sa_share_t sa_get_resource_parent(sa_resource_t);
+extern sa_resource_t sa_get_share_resource(sa_share_t, char *);
+extern sa_resource_t sa_add_resource(sa_share_t, char *, int, int *);
+extern int sa_remove_resource(sa_resource_t);
+extern char *sa_get_resource_attr(sa_resource_t, char *);
+extern int sa_set_resource_attr(sa_resource_t, char *, char *);
+extern int sa_set_resource_description(sa_resource_t, char *);
+extern char *sa_get_resource_description(sa_resource_t);
+extern int sa_enable_resource(sa_resource_t, char *);
+extern int sa_disable_resource(sa_resource_t, char *);
+extern int sa_rename_resource(sa_resource_t, char *);
+extern void sa_fix_resource_name(char *);
+
+/* data structure free calls */
+extern void sa_free_attr_string(char *);
+extern void sa_free_share_description(char *);
+
+/* optionset control */
+extern sa_optionset_t sa_get_optionset(sa_group_t, char *);
+extern sa_optionset_t sa_get_next_optionset(sa_group_t);
+extern char *sa_get_optionset_attr(sa_optionset_t, char *);
+extern void sa_set_optionset_attr(sa_optionset_t, char *, char *);
+extern sa_optionset_t sa_create_optionset(sa_group_t, char *);
+extern int sa_destroy_optionset(sa_optionset_t);
+extern sa_optionset_t sa_get_derived_optionset(void *, char *, int);
+extern void sa_free_derived_optionset(sa_optionset_t);
+
+/* property functions */
+extern sa_property_t sa_get_property(sa_optionset_t, char *);
+extern sa_property_t sa_get_next_property(sa_group_t);
+extern char *sa_get_property_attr(sa_property_t, char *);
+extern sa_property_t sa_create_section(char *, char *);
+extern void sa_set_section_attr(sa_property_t, char *, char *);
+extern sa_property_t sa_create_property(char *, char *);
+extern int sa_add_property(void *, sa_property_t);
+extern int sa_update_property(sa_property_t, char *);
+extern int sa_remove_property(sa_property_t);
+extern int sa_commit_properties(sa_optionset_t, int);
+extern int sa_valid_property(void *, char *, sa_property_t);
+extern int sa_is_persistent(void *);
+
+/* security control */
+extern sa_security_t sa_get_security(sa_group_t, char *, char *);
+extern sa_security_t sa_get_next_security(sa_security_t);
+extern char *sa_get_security_attr(sa_optionset_t, char *);
+extern sa_security_t sa_create_security(sa_group_t, char *, char *);
+extern int sa_destroy_security(sa_security_t);
+extern void sa_set_security_attr(sa_security_t, char *, char *);
+extern sa_optionset_t sa_get_all_security_types(void *, char *, int);
+extern sa_security_t sa_get_derived_security(void *, char *, char *, int);
+extern void sa_free_derived_security(sa_security_t);
+
+/* protocol specific interfaces */
+extern int sa_parse_legacy_options(sa_group_t, char *, char *);
+extern char *sa_proto_legacy_format(char *, sa_group_t, int);
+extern int sa_is_security(char *, char *);
+extern sa_protocol_properties_t sa_proto_get_properties(char *);
+extern uint64_t sa_proto_get_featureset(char *);
+extern sa_property_t sa_get_protocol_section(sa_protocol_properties_t, char *);
+extern sa_property_t sa_get_next_protocol_section(sa_property_t, char *);
+extern sa_property_t sa_get_protocol_property(sa_protocol_properties_t, char *);
+extern sa_property_t sa_get_next_protocol_property(sa_property_t, char *);
+extern int sa_set_protocol_property(sa_property_t, char *, char *);
+extern char *sa_get_protocol_status(char *);
+extern void sa_format_free(char *);
+extern sa_protocol_properties_t sa_create_protocol_properties(char *);
+extern int sa_add_protocol_property(sa_protocol_properties_t, sa_property_t);
+extern int sa_proto_valid_prop(char *, sa_property_t, sa_optionset_t);
+extern int sa_proto_valid_space(char *, char *);
+extern char *sa_proto_space_alias(char *, char *);
+extern int sa_proto_get_transients(sa_handle_t, char *);
+extern int sa_proto_notify_resource(sa_resource_t, char *);
+extern int sa_proto_change_notify(sa_share_t, char *);
+extern int sa_proto_delete_section(char *, char *);
+
+/* handle legacy (dfstab/sharetab) files */
+extern int sa_delete_legacy(sa_share_t, char *);
+extern int sa_update_legacy(sa_share_t, char *);
+extern int sa_update_sharetab(sa_share_t, char *);
+extern int sa_delete_sharetab(sa_handle_t, char *, char *);
+
+/* ZFS functions */
+extern int sa_zfs_is_shared(sa_handle_t, char *);
+extern int sa_group_is_zfs(sa_group_t);
+extern int sa_path_is_zfs(char *);
+
+/* SA Handle specific functions */
+extern sa_handle_t sa_find_group_handle(sa_group_t);
+
+#ifdef	__cplusplus
+}
+#endif
+
+#endif /* _LIBSHARE_H */
+#endif /* HAVE_LIBSHARE */
--- a/zfs/lib/libport/include/mntent.h
+++ b/zfs/lib/libport/include/mntent.h
@ -0,0 +1,35 @@
+/*
+ * CDDL HEADER START
+ *
+ * The contents of this file are subject to the terms of the
+ * Common Development and Distribution License, Version 1.0 only
+ * (the "License").  You may not use this file except in compliance
+ * with the License.
+ *
+ * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
+ * or http://www.opensolaris.org/os/licensing.
+ * See the License for the specific language governing permissions
+ * and limitations under the License.
+ *
+ * When distributing Covered Code, include this CDDL HEADER in each
+ * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
+ * If applicable, add the following below this CDDL HEADER, with the
+ * fields enclosed by brackets "[]" replaced with your own identifying
+ * information: Portions Copyright [yyyy] [name of copyright owner]
+ *
+ * CDDL HEADER END
+ */
+/*
+ * Copyright 2007 Sun Microsystems, Inc.  All rights reserved.
+ * Use is subject to license terms.
+ */
+
+#include_next <mntent.h>
+
+#ifndef _PORT_MNTENT_H
+#define _PORT_MNTENT_H
+
+/* For HAVE_SETMNTENT */
+#include "zfs_config.h"
+
+#endif
--- a/zfs/lib/libport/include/stdlib.h
+++ b/zfs/lib/libport/include/stdlib.h
@ -0,0 +1,38 @@
+/*
+ * CDDL HEADER START
+ *
+ * The contents of this file are subject to the terms of the
+ * Common Development and Distribution License, Version 1.0 only
+ * (the "License").  You may not use this file except in compliance
+ * with the License.
+ *
+ * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
+ * or http://www.opensolaris.org/os/licensing.
+ * See the License for the specific language governing permissions
+ * and limitations under the License.
+ *
+ * When distributing Covered Code, include this CDDL HEADER in each
+ * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
+ * If applicable, add the following below this CDDL HEADER, with the
+ * fields enclosed by brackets "[]" replaced with your own identifying
+ * information: Portions Copyright [yyyy] [name of copyright owner]
+ *
+ * CDDL HEADER END
+ */
+/*
+ * Copyright 2007 Sun Microsystems, Inc.  All rights reserved.
+ * Use is subject to license terms.
+ */
+
+#include_next <stdlib.h>
+
+#ifndef _PORT_STDLIB_H
+#define _PORT_STDLIB_H
+
+#include "zfs_config.h"
+
+#ifndef HAVE_GETEXECNAME
+extern const char *getexecname();
+#endif
+
+#endif
--- a/zfs/lib/libport/include/string.h
+++ b/zfs/lib/libport/include/string.h
@ -0,0 +1,46 @@
+/*
+ * CDDL HEADER START
+ *
+ * The contents of this file are subject to the terms of the
+ * Common Development and Distribution License, Version 1.0 only
+ * (the "License").  You may not use this file except in compliance
+ * with the License.
+ *
+ * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
+ * or http://www.opensolaris.org/os/licensing.
+ * See the License for the specific language governing permissions
+ * and limitations under the License.
+ *
+ * When distributing Covered Code, include this CDDL HEADER in each
+ * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
+ * If applicable, add the following below this CDDL HEADER, with the
+ * fields enclosed by brackets "[]" replaced with your own identifying
+ * information: Portions Copyright [yyyy] [name of copyright owner]
+ *
+ * CDDL HEADER END
+ */
+/*
+ * Copyright 2007 Sun Microsystems, Inc.  All rights reserved.
+ * Use is subject to license terms.
+ */
+
+#include_next <string.h>
+
+#ifndef _PORT_STRING_H
+#define _PORT_STRING_H
+
+#include "zfs_config.h"
+
+#ifndef HAVE_STRLCPY
+extern size_t strlcpy(char *dst, const char *src, size_t len);
+#endif
+
+#ifndef HAVE_STRLCAT
+extern size_t strlcat(char *, const char *, size_t);
+#endif
+
+#ifndef HAVE_STRNLEN
+extern size_t strnlen(const char *src, size_t maxlen);
+#endif
+
+#endif
--- a/zfs/lib/libport/include/strings.h
+++ b/zfs/lib/libport/include/strings.h
@ -0,0 +1,38 @@
+/*
+ * CDDL HEADER START
+ *
+ * The contents of this file are subject to the terms of the
+ * Common Development and Distribution License, Version 1.0 only
+ * (the "License").  You may not use this file except in compliance
+ * with the License.
+ *
+ * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
+ * or http://www.opensolaris.org/os/licensing.
+ * See the License for the specific language governing permissions
+ * and limitations under the License.
+ *
+ * When distributing Covered Code, include this CDDL HEADER in each
+ * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
+ * If applicable, add the following below this CDDL HEADER, with the
+ * fields enclosed by brackets "[]" replaced with your own identifying
+ * information: Portions Copyright [yyyy] [name of copyright owner]
+ *
+ * CDDL HEADER END
+ */
+/*
+ * Copyright 2007 Sun Microsystems, Inc.  All rights reserved.
+ * Use is subject to license terms.
+ */
+
+#include_next <strings.h>
+
+#ifndef _PORT_STRINGS_H
+#define _PORT_STRINGS_H
+
+#include "zfs_config.h"
+
+#ifndef HAVE_STRCMP_IN_STRINGS_H
+#include <string.h>
+#endif
+
+#endif
--- a/zfs/lib/libport/include/stropts.h
+++ b/zfs/lib/libport/include/stropts.h
@ -0,0 +1,37 @@
+/*
+ * CDDL HEADER START
+ *
+ * The contents of this file are subject to the terms of the
+ * Common Development and Distribution License (the "License").
+ * You may not use this file except in compliance with the License.
+ *
+ * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
+ * or http://www.opensolaris.org/os/licensing.
+ * See the License for the specific language governing permissions
+ * and limitations under the License.
+ *
+ * When distributing Covered Code, include this CDDL HEADER in each
+ * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
+ * If applicable, add the following below this CDDL HEADER, with the
+ * fields enclosed by brackets "[]" replaced with your own identifying
+ * information: Portions Copyright [yyyy] [name of copyright owner]
+ *
+ * CDDL HEADER END
+ */
+/*
+ * Copyright 2008 Sun Microsystems, Inc.  All rights reserved.
+ * Use is subject to license terms.
+ */
+
+#include_next <stropts.h>
+
+#ifndef _PORT_STROPTS_H
+#define _PORT_STROPTS_H
+
+#include "zfs_config.h"
+
+#ifdef HAVE_IOCTL_IN_STROPTS_H
+#include <fake_ioctl.h>
+#endif
+
+#endif /* _PORT_STROPTS_H */
--- a/Show More
+++ b/Show More