Tag zfs-2.0.6

META file and changelog updated.

Signed-off-by: Tony Hutter <hutter2@llnl.gov>

.github/ISSUE_TEMPLATE/bug_report.md

@@ -25,16 +25,14 @@ Type | Version/Name
  --- | ---
 Distribution Name	|
 Distribution Version	|
-Kernel Version	|
+Linux Kernel	|
 Architecture	|
-OpenZFS Version	|
+ZFS Version	|
+SPL Version	|
 <!--
-Command to find OpenZFS version:
+Commands to find ZFS/SPL versions:
-zfs version
+modinfo zfs | grep -iw version
-
+modinfo spl | grep -iw version
-Commands to find kernel version:
-uname -r		# Linux
-freebsd-version -r	# FreeBSD
 -->
 
 ### Describe the problem you're observing

.github/ISSUE_TEMPLATE/config.yml

@@ -1,8 +1,5 @@
 blank_issues_enabled: false
 contact_links:
-  - name: OpenZFS Questions
-    url: https://github.com/openzfs/zfs/discussions/new
-    about: Ask the community for help
   - name: OpenZFS Community Support Mailing list (Linux)
     url: https://zfsonlinux.topicbox.com/groups/zfs-discuss
     about: Get community support for OpenZFS on Linux
@@ -10,5 +7,5 @@ contact_links:
     url: https://lists.freebsd.org/mailman/listinfo/freebsd-fs
     about: Get community support for OpenZFS on FreeBSD
   - name: OpenZFS on IRC
-    url: https://web.libera.chat/#openzfs
+    url: https://webchat.freenode.net/#openzfs
     about: Use IRC to get community support for OpenZFS

.github/ISSUE_TEMPLATE/question.md

@@ -0,0 +1,37 @@
+---
+name: Code Question
+about: Ask a question about the code
+title: ''
+labels: 'Type: Question'
+assignees: ''
+
+---
+
+<!--
+Thank you for taking an interest in the OpenZFS codebase.
+
+Please be aware that most questions are preferably asked in the mailing list first.
+This form is primarily meant for asking questions about the code itself.
+
+Please also check our issue tracker before opening a new question.
+Filling out the following template will help other contributors better understand your question.
+-->
+
+### Ask your question!
+
+<!--
+Please provide a clear and concise question.
+-->
+
+### Which portion of the codebase does your question involve?
+
+<!--
+Optional: Please describe what portion of the codebase your issue involved.
+Example: "Testsuite", "Buildbots", "CLI", a code snippet etc.
+-->
+
+### Additional context
+
+<!--
+Any additional information you want to add?
+-->

.github/PULL_REQUEST_TEMPLATE.md

@@ -28,15 +28,14 @@ https://openzfs.github.io/openzfs-docs/Developer%20Resources/Buildbot%20Options.
 - [ ] Performance enhancement (non-breaking change which improves efficiency)
 - [ ] Code cleanup (non-breaking change which makes code smaller or more readable)
 - [ ] Breaking change (fix or feature that would cause existing functionality to change)
-- [ ] Library ABI change (libzfs, libzfs\_core, libnvpair, libuutil and libzfsbootenv)
 - [ ] Documentation (a change to man pages or other documentation)
 
 ### Checklist:
 <!--- Go over all the following points, and put an `x` in all the boxes that apply. -->
 <!--- If you're unsure about any of these, don't hesitate to ask. We're here to help! -->
-- [ ] My code follows the OpenZFS [code style requirements](https://github.com/openzfs/zfs/blob/master/.github/CONTRIBUTING.md#coding-conventions).
+- [ ] My code follows the ZFS on Linux [code style requirements](https://github.com/zfsonlinux/zfs/blob/master/.github/CONTRIBUTING.md#coding-conventions).
 - [ ] I have updated the documentation accordingly.
-- [ ] I have read the [**contributing** document](https://github.com/openzfs/zfs/blob/master/.github/CONTRIBUTING.md).
+- [ ] I have read the [**contributing** document](https://github.com/zfsonlinux/zfs/blob/master/.github/CONTRIBUTING.md).
-- [ ] I have added [tests](https://github.com/openzfs/zfs/tree/master/tests) to cover my changes.
+- [ ] I have added [tests](https://github.com/zfsonlinux/zfs/tree/master/tests) to cover my changes.
 - [ ] I have run the ZFS Test Suite with this change applied.
-- [ ] All commit messages are properly formatted and contain [`Signed-off-by`](https://github.com/openzfs/zfs/blob/master/.github/CONTRIBUTING.md#signed-off-by).
+- [ ] All commit messages are properly formatted and contain [`Signed-off-by`](https://github.com/zfsonlinux/zfs/blob/master/.github/CONTRIBUTING.md#signed-off-by).

.github/no-response.yml

@@ -1,13 +0,0 @@
-# Configuration for probot-no-response - https://github.com/probot/no-response
-
-# Number of days of inactivity before an Issue is closed for lack of response
-daysUntilClose: 31
-# Label requiring a response
-responseRequiredLabel: "Status: Feedback requested"
-# Comment to post when closing an Issue for lack of response. Set to `false` to disable
-closeComment: >
-  This issue has been automatically closed because there has been no response
-  to our request for more information from the original author. With only the
-  information that is currently in the issue, we don't have enough information
-  to take action. Please reach out if you have or find the answers we need so
-  that we can investigate further.

.github/stale.yml

@@ -7,14 +7,7 @@ only: issues
 # Issues with these labels will never be considered stale
 exemptLabels:
   - "Type: Feature"
-  - "Bot: Not Stale"
+  - "Type: Understood"
-  - "Status: Work in Progress"
-# Set to true to ignore issues in a project (defaults to false)
-exemptProjects: true
-# Set to true to ignore issues in a milestone (defaults to false)
-exemptMilestones: true
-# Set to true to ignore issues with an assignee (defaults to false)
-exemptAssignees: true
 # Label to use when marking an issue as stale
 staleLabel: "Status: Stale"
 # Comment to post when marking an issue as stale. Set to `false` to disable
@@ -22,5 +15,3 @@ markComment: >
   This issue has been automatically marked as "stale" because it has not had
   any activity for a while. It will be closed in 90 days if no further activity occurs. 
   Thank you for your contributions.
-# Limit the number of actions per hour, from 1-30. Default is 30
-limitPerRun: 6

.github/workflows/checkstyle.yaml

@@ -6,7 +6,7 @@ on:
 
 jobs:
   checkstyle:
-    runs-on: ubuntu-20.04
+    runs-on: ubuntu-18.04
     steps:
     - uses: actions/checkout@v2
       with:
@@ -18,13 +18,12 @@ jobs:
         sudo apt-get install --yes -qq zlib1g-dev uuid-dev libattr1-dev libblkid-dev libselinux-dev libudev-dev libssl-dev python-dev python-setuptools python-cffi python3 python3-dev python3-setuptools python3-cffi
         # packages for tests
         sudo apt-get install --yes -qq parted lsscsi ksh attr acl nfs-kernel-server fio
-        sudo apt-get install --yes -qq mandoc cppcheck pax-utils devscripts
+        sudo apt-get install --yes -qq mandoc cppcheck pax-utils abigail-tools # devscripts - enable then bashisms fixed
         sudo -E pip --quiet install flake8
     - name: Prepare
       run: |
         sh ./autogen.sh
         ./configure
-        make -j$(nproc)
     - name: Checkstyle
       run: |
         make checkstyle
@@ -32,19 +31,6 @@ jobs:
       run: |
         make lint
     - name: CheckABI
-      id: CheckABI
       run: |
-        sudo docker run -v $(pwd):/source ghcr.io/openzfs/libabigail make checkabi
+        make -j$(nproc)
-    - name: StoreABI
+        make checkabi
-      if: failure() && steps.CheckABI.outcome == 'failure'
-      run: |
-        sudo docker run -v $(pwd):/source ghcr.io/openzfs/libabigail make storeabi
-    - name: Prepare artifacts
-      if: failure() && steps.CheckABI.outcome == 'failure'
-      run: |
-        find -name *.abi | tar -cf abi_files.tar -T -
-    - uses: actions/upload-artifact@v2
-      if: failure() && steps.CheckABI.outcome == 'failure'
-      with:
-        name: New ABI files (use only if you're sure about interface changes)
-        path: abi_files.tar

.github/workflows/zfs-tests-functional.yml

@@ -26,8 +26,7 @@ jobs:
           xfslibs-dev libattr1-dev libacl1-dev libudev-dev libdevmapper-dev \
           libssl-dev libffi-dev libaio-dev libelf-dev libmount-dev \
           libpam0g-dev pamtester python-dev python-setuptools python-cffi \
-          python3 python3-dev python3-setuptools python3-cffi python3-packaging \
+          python3 python3-dev python3-setuptools python3-cffi
-          libcurl4-openssl-dev
     - name: Autogen.sh
       run: |
         sh autogen.sh
@@ -45,17 +44,6 @@ jobs:
         sudo sed -i.bak 's/updates/extra updates/' /etc/depmod.d/ubuntu.conf
         sudo depmod
         sudo modprobe zfs
-        # Workaround for cloud-init bug
-        # see https://github.com/openzfs/zfs/issues/12644
-        FILE=/lib/udev/rules.d/10-cloud-init-hook-hotplug.rules
-        if [ -r "${FILE}" ]; then
-          HASH=$(md5sum "${FILE}" | awk '{ print $1 }')
-          if [ "${HASH}" = "121ff0ef1936cd2ef65aec0458a35772" ]; then
-            # Just shove a zd* exclusion right above the hotplug hook...
-            sudo sed -i -e s/'LABEL="cloudinit_hook"'/'KERNEL=="zd*", GOTO="cloudinit_end"\n&'/ "${FILE}"
-            sudo udevadm control --reload-rules
-          fi
-        fi
         # Workaround to provide additional free space for testing.
         #   https://github.com/actions/virtual-environments/issues/2840
         sudo rm -rf /usr/share/dotnet
@@ -64,8 +52,7 @@ jobs:
         sudo rm -rf "$AGENT_TOOLSDIRECTORY"
     - name: Tests
       run: |
-        /usr/share/zfs/zfs-tests.sh -vR -s 3G
+        /usr/share/zfs/zfs-tests.sh -v -s 3G
-      timeout-minutes: 330
     - name: Prepare artifacts
       if: failure()
       run: |
@@ -74,7 +61,7 @@ jobs:
         sudo cp /var/log/syslog $RESULTS_PATH/
         sudo chmod +r $RESULTS_PATH/*
         # Replace ':' in dir names, actions/upload-artifact doesn't support it
-        for f in $(find /var/tmp/test_results -name '*:*'); do mv "$f" "${f//:/__}"; done
+        for f in $(find $RESULTS_PATH -name '*:*'); do mv "$f" "${f//:/__}"; done
     - uses: actions/upload-artifact@v2
       if: failure()
       with:

.github/workflows/zfs-tests-sanity.yml

@@ -22,8 +22,7 @@ jobs:
           xfslibs-dev libattr1-dev libacl1-dev libudev-dev libdevmapper-dev \
           libssl-dev libffi-dev libaio-dev libelf-dev libmount-dev \
           libpam0g-dev pamtester python-dev python-setuptools python-cffi \
-          python3 python3-dev python3-setuptools python3-cffi python3-packaging \
+          python3 python3-dev python3-setuptools python3-cffi
-          libcurl4-openssl-dev
     - name: Autogen.sh
       run: |
         sh autogen.sh
@@ -41,17 +40,6 @@ jobs:
         sudo sed -i.bak 's/updates/extra updates/' /etc/depmod.d/ubuntu.conf
         sudo depmod
         sudo modprobe zfs
-        # Workaround for cloud-init bug
-        # see https://github.com/openzfs/zfs/issues/12644
-        FILE=/lib/udev/rules.d/10-cloud-init-hook-hotplug.rules
-        if [ -r "${FILE}" ]; then
-          HASH=$(md5sum "${FILE}" | awk '{ print $1 }')
-          if [ "${HASH}" = "121ff0ef1936cd2ef65aec0458a35772" ]; then
-            # Just shove a zd* exclusion right above the hotplug hook...
-            sudo sed -i -e s/'LABEL="cloudinit_hook"'/'KERNEL=="zd*", GOTO="cloudinit_end"\n&'/ "${FILE}"
-            sudo udevadm control --reload-rules
-          fi
-        fi
         # Workaround to provide additional free space for testing.
         #   https://github.com/actions/virtual-environments/issues/2840
         sudo rm -rf /usr/share/dotnet
@@ -60,8 +48,7 @@ jobs:
         sudo rm -rf "$AGENT_TOOLSDIRECTORY"
     - name: Tests
       run: |
-        /usr/share/zfs/zfs-tests.sh -vR -s 3G -r sanity
+        /usr/share/zfs/zfs-tests.sh -v -s 3G -r sanity
-      timeout-minutes: 330
     - name: Prepare artifacts
       if: failure()
       run: |
@@ -70,7 +57,7 @@ jobs:
         sudo cp /var/log/syslog $RESULTS_PATH/
         sudo chmod +r $RESULTS_PATH/*
         # Replace ':' in dir names, actions/upload-artifact doesn't support it
-        for f in $(find /var/tmp/test_results -name '*:*'); do mv "$f" "${f//:/__}"; done
+        for f in $(find $RESULTS_PATH -name '*:*'); do mv "$f" "${f//:/__}"; done
     - uses: actions/upload-artifact@v2
       if: failure()
       with:

.github/workflows/zloop.yml

@@ -22,8 +22,8 @@ jobs:
           xfslibs-dev libattr1-dev libacl1-dev libudev-dev libdevmapper-dev \
           libssl-dev libffi-dev libaio-dev libelf-dev libmount-dev \
           libpam0g-dev \
-          python-dev python-setuptools python-cffi python-packaging \
+          python-dev python-setuptools python-cffi \
-          python3 python3-dev python3-setuptools python3-cffi python3-packaging
+          python3 python3-dev python3-setuptools python3-cffi
     - name: Autogen.sh
       run: |
         sh autogen.sh
@@ -45,7 +45,7 @@ jobs:
       run: |
         sudo mkdir -p $TEST_DIR
         # run for 20 minutes to have a total runner time of 30 minutes
-        sudo /usr/share/zfs/zloop.sh -t 1200 -l -m1 -- -T 120 -P 60
+        sudo /usr/share/zfs/zloop.sh -t 1200 -l -m1
     - name: Prepare artifacts
       if: failure()
       run: |

.gitmodules

@@ -1,3 +1,3 @@
 [submodule "scripts/zfs-images"]
 	path = scripts/zfs-images
-	url = https://github.com/openzfs/zfs-images
+	url = https://github.com/zfsonlinux/zfs-images

META

@@ -1,10 +1,10 @@
 Meta:          1
 Name:          zfs
 Branch:        1.0
-Version:       2.1.5
+Version:       2.0.6
 Release:       1
 Release-Tags:  relext
 License:       CDDL
 Author:        OpenZFS
-Linux-Maximum: 5.18
+Linux-Maximum: 5.14
 Linux-Minimum: 3.10

Makefile.am

@@ -1,5 +1,3 @@
-include $(top_srcdir)/config/Shellcheck.am
-
 ACLOCAL_AMFLAGS = -I config
 
 SUBDIRS = include
@@ -8,7 +6,7 @@ SUBDIRS += rpm
 endif
 
 if CONFIG_USER
-SUBDIRS += man scripts lib tests cmd etc contrib
+SUBDIRS += etc man scripts lib tests cmd contrib
 if BUILD_LINUX
 SUBDIRS += udev
 endif
@@ -28,8 +26,8 @@ endif
 AUTOMAKE_OPTIONS = foreign
 EXTRA_DIST  = autogen.sh copy-builtin
 EXTRA_DIST += config/config.awk config/rpm.am config/deb.am config/tgz.am
-EXTRA_DIST += AUTHORS CODE_OF_CONDUCT.md COPYRIGHT LICENSE META NEWS NOTICE
+EXTRA_DIST += META AUTHORS COPYRIGHT LICENSE NEWS NOTICE README.md
-EXTRA_DIST += README.md RELEASES.md
+EXTRA_DIST += CODE_OF_CONDUCT.md
 EXTRA_DIST += module/lua/README.zfs module/os/linux/spl/README.md
 
 # Include all the extra licensing information for modules
@@ -103,7 +101,7 @@ endif
 endif
 
 PHONY += codecheck
-codecheck: cstyle shellcheck checkbashisms flake8 mancheck testscheck vcscheck zstdcheck
+codecheck: cstyle shellcheck checkbashisms flake8 mancheck testscheck vcscheck
 
 PHONY += checkstyle
 checkstyle: codecheck commitcheck
@@ -120,36 +118,64 @@ cstyle:
 		-o -type f -name '*.[hc]' \
 		! -name 'zfs_config.*' ! -name '*.mod.c' \
 		! -name 'opt_global.h' ! -name '*_if*.h' \
-		! -name 'zstd_compat_wrapper.h' \
 		! -path './module/zstd/lib/*' \
 		-exec ${top_srcdir}/scripts/cstyle.pl -cpP {} \+
 
 filter_executable = -exec test -x '{}' \; -print
 
-SHELLCHECKDIRS = cmd contrib etc scripts tests
+PHONY += shellcheck
-SHELLCHECKSCRIPTS = autogen.sh
+shellcheck:
+	@if type shellcheck > /dev/null 2>&1; then \
+		shellcheck --exclude=SC1090 --exclude=SC1117 --format=gcc \
+			$$(find ${top_srcdir}/scripts/*.sh -type f) \
+			$$(find ${top_srcdir}/cmd/zed/zed.d/*.sh -type f) \
+			$$(find ${top_srcdir}/cmd/zpool/zpool.d/* \
+			-type f ${filter_executable}); \
+	else \
+		echo "skipping shellcheck because shellcheck is not installed"; \
+	fi
 
 PHONY += checkabi storeabi
-
+checkabi: lib
-checklibabiversion:
-	libabiversion=`abidw -v | $(SED) 's/[^0-9]//g'`; \
-	if test $$libabiversion -lt "200"; then \
-        /bin/echo -e "\n" \
-        "*** Please use libabigail 2.0.0 version or newer;\n" \
-        "*** otherwise results are not consistent!\n" \
-        "(or see https://github.com/openzfs/libabigail-docker )\n"; \
-        exit 1; \
-    fi;
-
-checkabi: checklibabiversion lib
 	$(MAKE) -C lib checkabi
 
-storeabi: checklibabiversion lib
+storeabi: lib
 	$(MAKE) -C lib storeabi
 
+PHONY += checkbashisms
+checkbashisms:
+	@if type checkbashisms > /dev/null 2>&1; then \
+		checkbashisms -n -p -x \
+			$$(find ${top_srcdir} \
+				-name '.git' -prune \
+				-o -name 'build' -prune \
+				-o -name 'tests' -prune \
+				-o -name 'config' -prune \
+				-o -name 'zed-functions.sh*' -prune \
+				-o -name 'zfs-import*' -prune \
+				-o -name 'zfs-mount*' -prune \
+				-o -name 'zfs-zed*' -prune \
+				-o -name 'smart' -prune \
+				-o -name 'paxcheck.sh' -prune \
+				-o -name 'make_gitrev.sh' -prune \
+				-o -name '90zfs' -prune \
+				-o -type f ! -name 'config*' \
+				! -name 'libtool' \
+			-exec sh -c 'awk "NR==1 && /#!.*bin\/sh.*/ {print FILENAME;}" "{}"' \;); \
+	else \
+		echo "skipping checkbashisms because checkbashisms is not installed"; \
+	fi
+
 PHONY += mancheck
 mancheck:
-	${top_srcdir}/scripts/mancheck.sh ${top_srcdir}/man ${top_srcdir}/tests/test-runner/man
+	@if type mandoc > /dev/null 2>&1; then \
+		find ${top_srcdir}/man/man8 -type f -name 'zfs.8' \
+			-o -name 'zpool.8' -o -name 'zdb.8' \
+			-o -name 'zgenhostid.8' | \
+			xargs mandoc -Tlint -Werror; \
+	else \
+		echo "skipping mancheck because mandoc is not installed"; \
+	fi
 
 if BUILD_LINUX
 stat_fmt = -c '%A %n'
@@ -174,10 +200,6 @@ vcscheck:
 		awk '{c++; print} END {if(c>0) exit 1}' ; \
 	fi
 
-PHONY += zstdcheck
-zstdcheck:
-	@$(MAKE) -C module/zstd checksymbols
-
 PHONY += lint
 lint: cppcheck paxcheck
 

README.md

@@ -12,7 +12,7 @@ This repository contains the code for running OpenZFS on Linux and FreeBSD.
   * [Documentation](https://openzfs.github.io/openzfs-docs/) - for using and developing this repo
   * [ZoL Site](https://zfsonlinux.org) - Linux release info & links
   * [Mailing lists](https://openzfs.github.io/openzfs-docs/Project%20and%20Community/Mailing%20Lists.html)
-  * [OpenZFS site](https://openzfs.org/) - for conference videos and info on other platforms (illumos, OSX, Windows, etc)
+  * [OpenZFS site](http://open-zfs.org/) - for conference videos and info on other platforms (illumos, OSX, Windows, etc)
 
 # Installation
 
@@ -32,4 +32,4 @@ For more details see the NOTICE, LICENSE and COPYRIGHT files; `UCRL-CODE-235197`
 
 # Supported Kernels
   * The `META` file contains the officially recognized supported Linux kernel versions.
-  * Supported FreeBSD versions are any supported branches and releases starting from 12.2-RELEASE.
+  * Supported FreeBSD versions are 12-STABLE and 13-CURRENT.

RELEASES.md

@@ -1,37 +0,0 @@
-OpenZFS uses the MAJOR.MINOR.PATCH versioning scheme described here:
-
-  * MAJOR - Incremented at the discretion of the OpenZFS developers to indicate
-    a particularly noteworthy feature or change. An increase in MAJOR number
-    does not indicate any incompatible on-disk format change. The ability
-    to import a ZFS pool is controlled by the feature flags enabled on the
-    pool and the feature flags supported by the installed OpenZFS version.
-    Increasing the MAJOR version is expected to be an infrequent occurrence.
-
-  * MINOR - Incremented to indicate new functionality such as a new feature
-    flag, pool/dataset property, zfs/zpool sub-command, new user/kernel
-    interface, etc. MINOR releases may introduce incompatible changes to the
-    user space library APIs (libzfs.so). Existing user/kernel interfaces are
-    considered to be stable to maximize compatibility between OpenZFS releases.
-    Additions to the user/kernel interface are backwards compatible.
-
-  * PATCH - Incremented when applying documentation updates, important bug
-    fixes, minor performance improvements, and kernel compatibility patches.
-    The user space library APIs and user/kernel interface are considered to
-    be stable. PATCH releases for a MAJOR.MINOR are published as needed.
-
-Two release branches are maintained for OpenZFS, they are:
-
-  * OpenZFS LTS - A designated MAJOR.MINOR release with periodic PATCH
-    releases that incorporate important changes backported from newer OpenZFS
-    releases. This branch is intended for use in environments using an
-    LTS, enterprise, or similarly managed kernel (RHEL, Ubuntu LTS, Debian).
-    Minor changes to support these distribution kernels will be applied as
-    needed. New kernel versions released after the OpenZFS LTS release are
-    not supported. LTS releases will receive patches for at least 2 years.
-    The current LTS release is OpenZFS 2.1.
-
-  * OpenZFS current - Tracks the newest MAJOR.MINOR release. This branch
-    includes support for the latest OpenZFS features and recently releases
-    kernels.  When a new MINOR release is tagged the previous MINOR release
-    will no longer be maintained (unless it is an LTS release). New MINOR
-    releases are planned to occur roughly annually.

cmd/Makefile.am

@@ -1,14 +1,8 @@
-include $(top_srcdir)/config/Shellcheck.am
+SUBDIRS  = zfs zpool zdb zhack zinject zstream zstreamdump ztest
-
-SUBDIRS  = zfs zpool zdb zhack zinject zstream ztest
 SUBDIRS += fsck_zfs vdev_id raidz_test zfs_ids_to_path
-SUBDIRS += zpool_influxdb
 
 CPPCHECKDIRS  = zfs zpool zdb zhack zinject zstream ztest
-CPPCHECKDIRS += raidz_test zfs_ids_to_path zpool_influxdb
+CPPCHECKDIRS += raidz_test zfs_ids_to_path
-
-# TODO: #12084: SHELLCHECKDIRS = fsck_zfs vdev_id zpool
-SHELLCHECKDIRS = fsck_zfs zpool
 
 if USING_PYTHON
 SUBDIRS += arcstat arc_summary dbufstat
@@ -17,7 +11,6 @@ endif
 if BUILD_LINUX
 SUBDIRS += mount_zfs zed zgenhostid zvol_id zvol_wait
 CPPCHECKDIRS += mount_zfs zed zgenhostid zvol_id
-SHELLCHECKDIRS += zed
 endif
 
 PHONY = cppcheck

cmd/arc_summary/arc_summary2

@@ -102,6 +102,18 @@ show_tunable_descriptions = False
 alternate_tunable_layout = False
 
 
+def handle_Exception(ex_cls, ex, tb):
+    if ex is IOError:
+        if ex.errno == errno.EPIPE:
+            sys.exit()
+
+    if ex is KeyboardInterrupt:
+        sys.exit()
+
+
+sys.excepthook = handle_Exception
+
+
 def get_Kstat():
     """Collect information on the ZFS subsystem from the /proc virtual
     file system. The name "kstat" is a holdover from the Solaris utility
@@ -213,30 +225,12 @@ def get_arc_summary(Kstat):
     deleted = Kstat["kstat.zfs.misc.arcstats.deleted"]
     mutex_miss = Kstat["kstat.zfs.misc.arcstats.mutex_miss"]
     evict_skip = Kstat["kstat.zfs.misc.arcstats.evict_skip"]
-    evict_l2_cached = Kstat["kstat.zfs.misc.arcstats.evict_l2_cached"]
-    evict_l2_eligible = Kstat["kstat.zfs.misc.arcstats.evict_l2_eligible"]
-    evict_l2_eligible_mfu = Kstat["kstat.zfs.misc.arcstats.evict_l2_eligible_mfu"]
-    evict_l2_eligible_mru = Kstat["kstat.zfs.misc.arcstats.evict_l2_eligible_mru"]
-    evict_l2_ineligible = Kstat["kstat.zfs.misc.arcstats.evict_l2_ineligible"]
-    evict_l2_skip = Kstat["kstat.zfs.misc.arcstats.evict_l2_skip"]
 
     # ARC Misc.
     output["arc_misc"] = {}
     output["arc_misc"]["deleted"] = fHits(deleted)
-    output["arc_misc"]["mutex_miss"] = fHits(mutex_miss)
+    output["arc_misc"]['mutex_miss'] = fHits(mutex_miss)
-    output["arc_misc"]["evict_skips"] = fHits(evict_skip)
+    output["arc_misc"]['evict_skips'] = fHits(evict_skip)
-    output["arc_misc"]["evict_l2_skip"] = fHits(evict_l2_skip)
-    output["arc_misc"]["evict_l2_cached"] = fBytes(evict_l2_cached)
-    output["arc_misc"]["evict_l2_eligible"] = fBytes(evict_l2_eligible)
-    output["arc_misc"]["evict_l2_eligible_mfu"] = {
-            'per': fPerc(evict_l2_eligible_mfu, evict_l2_eligible),
-            'num': fBytes(evict_l2_eligible_mfu),
-    }
-    output["arc_misc"]["evict_l2_eligible_mru"] = {
-            'per': fPerc(evict_l2_eligible_mru, evict_l2_eligible),
-            'num': fBytes(evict_l2_eligible_mru),
-    }
-    output["arc_misc"]["evict_l2_ineligible"] = fBytes(evict_l2_ineligible)
 
     # ARC Sizing
     arc_size = Kstat["kstat.zfs.misc.arcstats.size"]
@@ -352,26 +346,8 @@ def _arc_summary(Kstat):
     sys.stdout.write("\tDeleted:\t\t\t\t%s\n" % arc['arc_misc']['deleted'])
     sys.stdout.write("\tMutex Misses:\t\t\t\t%s\n" %
                      arc['arc_misc']['mutex_miss'])
-    sys.stdout.write("\tEviction Skips:\t\t\t\t%s\n" %
+    sys.stdout.write("\tEvict Skips:\t\t\t\t%s\n" %
                      arc['arc_misc']['evict_skips'])
-    sys.stdout.write("\tEviction Skips Due to L2 Writes:\t%s\n" %
-                     arc['arc_misc']['evict_l2_skip'])
-    sys.stdout.write("\tL2 Cached Evictions:\t\t\t%s\n" %
-                     arc['arc_misc']['evict_l2_cached'])
-    sys.stdout.write("\tL2 Eligible Evictions:\t\t\t%s\n" %
-                     arc['arc_misc']['evict_l2_eligible'])
-    sys.stdout.write("\tL2 Eligible MFU Evictions:\t%s\t%s\n" % (
-                     arc['arc_misc']['evict_l2_eligible_mfu']['per'],
-                     arc['arc_misc']['evict_l2_eligible_mfu']['num'],
-                     )
-    )
-    sys.stdout.write("\tL2 Eligible MRU Evictions:\t%s\t%s\n" % (
-                     arc['arc_misc']['evict_l2_eligible_mru']['per'],
-                     arc['arc_misc']['evict_l2_eligible_mru']['num'],
-                     )
-    )
-    sys.stdout.write("\tL2 Ineligible Evictions:\t\t%s\n" %
-                     arc['arc_misc']['evict_l2_ineligible'])
     sys.stdout.write("\n")
 
     # ARC Sizing
@@ -707,11 +683,6 @@ def get_l2arc_summary(Kstat):
     l2_writes_done = Kstat["kstat.zfs.misc.arcstats.l2_writes_done"]
     l2_writes_error = Kstat["kstat.zfs.misc.arcstats.l2_writes_error"]
     l2_writes_sent = Kstat["kstat.zfs.misc.arcstats.l2_writes_sent"]
-    l2_mfu_asize = Kstat["kstat.zfs.misc.arcstats.l2_mfu_asize"]
-    l2_mru_asize = Kstat["kstat.zfs.misc.arcstats.l2_mru_asize"]
-    l2_prefetch_asize = Kstat["kstat.zfs.misc.arcstats.l2_prefetch_asize"]
-    l2_bufc_data_asize = Kstat["kstat.zfs.misc.arcstats.l2_bufc_data_asize"]
-    l2_bufc_metadata_asize = Kstat["kstat.zfs.misc.arcstats.l2_bufc_metadata_asize"]
 
     l2_access_total = (l2_hits + l2_misses)
     output['l2_health_count'] = (l2_writes_error + l2_cksum_bad + l2_io_error)
@@ -734,7 +705,7 @@ def get_l2arc_summary(Kstat):
         output["io_errors"] = fHits(l2_io_error)
 
         output["l2_arc_size"] = {}
-        output["l2_arc_size"]["adaptive"] = fBytes(l2_size)
+        output["l2_arc_size"]["adative"] = fBytes(l2_size)
         output["l2_arc_size"]["actual"] = {
             'per': fPerc(l2_asize, l2_size),
             'num': fBytes(l2_asize)
@@ -743,26 +714,6 @@ def get_l2arc_summary(Kstat):
             'per': fPerc(l2_hdr_size, l2_size),
             'num': fBytes(l2_hdr_size),
         }
-        output["l2_arc_size"]["mfu_asize"] = {
-            'per': fPerc(l2_mfu_asize, l2_asize),
-            'num': fBytes(l2_mfu_asize),
-        }
-        output["l2_arc_size"]["mru_asize"] = {
-            'per': fPerc(l2_mru_asize, l2_asize),
-            'num': fBytes(l2_mru_asize),
-        }
-        output["l2_arc_size"]["prefetch_asize"] = {
-            'per': fPerc(l2_prefetch_asize, l2_asize),
-            'num': fBytes(l2_prefetch_asize),
-        }
-        output["l2_arc_size"]["bufc_data_asize"] = {
-            'per': fPerc(l2_bufc_data_asize, l2_asize),
-            'num': fBytes(l2_bufc_data_asize),
-        }
-        output["l2_arc_size"]["bufc_metadata_asize"] = {
-            'per': fPerc(l2_bufc_metadata_asize, l2_asize),
-            'num': fBytes(l2_bufc_metadata_asize),
-        }
 
         output["l2_arc_evicts"] = {}
         output["l2_arc_evicts"]['lock_retries'] = fHits(l2_evict_lock_retry)
@@ -827,7 +778,7 @@ def _l2arc_summary(Kstat):
         sys.stdout.write("\n")
 
         sys.stdout.write("L2 ARC Size: (Adaptive)\t\t\t\t%s\n" %
-                         arc["l2_arc_size"]["adaptive"])
+                         arc["l2_arc_size"]["adative"])
         sys.stdout.write("\tCompressed:\t\t\t%s\t%s\n" % (
             arc["l2_arc_size"]["actual"]["per"],
             arc["l2_arc_size"]["actual"]["num"],
@@ -838,36 +789,11 @@ def _l2arc_summary(Kstat):
             arc["l2_arc_size"]["head_size"]["num"],
             )
         )
-        sys.stdout.write("\tMFU Alloc. Size:\t\t%s\t%s\n" % (
-            arc["l2_arc_size"]["mfu_asize"]["per"],
-            arc["l2_arc_size"]["mfu_asize"]["num"],
-            )
-        )
-        sys.stdout.write("\tMRU Alloc. Size:\t\t%s\t%s\n" % (
-            arc["l2_arc_size"]["mru_asize"]["per"],
-            arc["l2_arc_size"]["mru_asize"]["num"],
-            )
-        )
-        sys.stdout.write("\tPrefetch Alloc. Size:\t\t%s\t%s\n" % (
-            arc["l2_arc_size"]["prefetch_asize"]["per"],
-            arc["l2_arc_size"]["prefetch_asize"]["num"],
-            )
-        )
-        sys.stdout.write("\tData (buf content) Alloc. Size:\t%s\t%s\n" % (
-            arc["l2_arc_size"]["bufc_data_asize"]["per"],
-            arc["l2_arc_size"]["bufc_data_asize"]["num"],
-            )
-        )
-        sys.stdout.write("\tMetadata (buf content) Size:\t%s\t%s\n" % (
-            arc["l2_arc_size"]["bufc_metadata_asize"]["per"],
-            arc["l2_arc_size"]["bufc_metadata_asize"]["num"],
-            )
-        )
         sys.stdout.write("\n")
 
         if arc["l2_arc_evicts"]['lock_retries'] != '0' or \
            arc["l2_arc_evicts"]["reading"] != '0':
-            sys.stdout.write("L2 ARC Evictions:\n")
+            sys.stdout.write("L2 ARC Evicts:\n")
             sys.stdout.write("\tLock Retries:\t\t\t\t%s\n" %
                              arc["l2_arc_evicts"]['lock_retries'])
             sys.stdout.write("\tUpon Reading:\t\t\t\t%s\n" %
@@ -1125,55 +1051,48 @@ def main():
     global alternate_tunable_layout
 
     try:
-        try:
+        opts, args = getopt.getopt(
-            opts, args = getopt.getopt(
+            sys.argv[1:],
-                sys.argv[1:],
+            "adp:h", ["alternate", "description", "page=", "help"]
-                "adp:h", ["alternate", "description", "page=", "help"]
+        )
-            )
+    except getopt.error as e:
-        except getopt.error as e:
+        sys.stderr.write("Error: %s\n" % e.msg)
-            sys.stderr.write("Error: %s\n" % e.msg)
+        usage()
+        sys.exit(1)
+
+    args = {}
+    for opt, arg in opts:
+        if opt in ('-a', '--alternate'):
+            args['a'] = True
+        if opt in ('-d', '--description'):
+            args['d'] = True
+        if opt in ('-p', '--page'):
+            args['p'] = arg
+        if opt in ('-h', '--help'):
             usage()
-            sys.exit(1)
-
-        args = {}
-        for opt, arg in opts:
-            if opt in ('-a', '--alternate'):
-                args['a'] = True
-            if opt in ('-d', '--description'):
-                args['d'] = True
-            if opt in ('-p', '--page'):
-                args['p'] = arg
-            if opt in ('-h', '--help'):
-                usage()
-                sys.exit(0)
-
-        Kstat = get_Kstat()
-
-        alternate_tunable_layout = 'a' in args
-        show_tunable_descriptions = 'd' in args
-
-        pages = []
-
-        if 'p' in args:
-            try:
-                pages.append(unSub[int(args['p']) - 1])
-            except IndexError:
-                sys.stderr.write('the argument to -p must be between 1 and ' +
-                                 str(len(unSub)) + '\n')
-                sys.exit(1)
-        else:
-            pages = unSub
-
-        zfs_header()
-        for page in pages:
-            page(Kstat)
-            sys.stdout.write("\n")
-    except IOError as ex:
-        if (ex.errno == errno.EPIPE):
             sys.exit(0)
-        raise
+
-    except KeyboardInterrupt:
+    Kstat = get_Kstat()
-        sys.exit(0)
+
+    alternate_tunable_layout = 'a' in args
+    show_tunable_descriptions = 'd' in args
+
+    pages = []
+
+    if 'p' in args:
+        try:
+            pages.append(unSub[int(args['p']) - 1])
+        except IndexError:
+            sys.stderr.write('the argument to -p must be between 1 and ' +
+                             str(len(unSub)) + '\n')
+            sys.exit(1)
+    else:
+        pages = unSub
+
+    zfs_header()
+    for page in pages:
+        page(Kstat)
+        sys.stdout.write("\n")
 
 
 if __name__ == '__main__':

cmd/arc_summary/arc_summary3

@@ -42,13 +42,6 @@ import os
 import subprocess
 import sys
 import time
-import errno
-
-# We can't use env -S portably, and we need python3 -u to handle pipes in
-# the shell abruptly closing the way we want to, so...
-import io
-if isinstance(sys.__stderr__.buffer, io.BufferedWriter):
-    os.execv(sys.executable, [sys.executable, "-u"] + sys.argv)
 
 DESCRIPTION = 'Print ARC and other statistics for OpenZFS'
 INDENT = ' '*8
@@ -65,6 +58,7 @@ SECTION_PATHS = {'arc': 'arcstats',
                  'dmu': 'dmu_tx',
                  'l2arc': 'arcstats',  # L2ARC stuff lives in arcstats
                  'vdev': 'vdev_cache_stats',
+                 'xuio': 'xuio_stats',
                  'zfetch': 'zfetchstats',
                  'zil': 'zil'}
 
@@ -168,11 +162,21 @@ elif sys.platform.startswith('linux'):
         # The original arc_summary called /sbin/modinfo/{spl,zfs} to get
         # the version information. We switch to /sys/module/{spl,zfs}/version
         # to make sure we get what is really loaded in the kernel
-        try:
+        command = ["cat", "/sys/module/{0}/version".format(request)]
-            with open("/sys/module/{}/version".format(request)) as f:
+        req = request.upper()
-                return f.read().strip()
+
-        except:
+        # The recommended way to do this is with subprocess.run(). However,
-            return "(unknown)"
+        # some installed versions of Python are < 3.5, so we offer them
+        # the option of doing it the old way (for now)
+        if 'run' in dir(subprocess):
+            info = subprocess.run(command, stdout=subprocess.PIPE,
+                                  universal_newlines=True)
+            version = info.stdout.strip()
+        else:
+            info = subprocess.check_output(command, universal_newlines=True)
+            version = info.strip()
+
+        return version
 
     def get_descriptions(request):
         """Get the descriptions of the Solaris Porting Layer (SPL) or the
@@ -228,29 +232,6 @@ elif sys.platform.startswith('linux'):
 
         return descs
 
-def handle_unraisableException(exc_type, exc_value=None, exc_traceback=None,
-                               err_msg=None, object=None):
-   handle_Exception(exc_type, object, exc_traceback)
-
-def handle_Exception(ex_cls, ex, tb):
-    if ex_cls is KeyboardInterrupt:
-        sys.exit()
-
-    if ex_cls is BrokenPipeError:
-        # It turns out that while sys.exit() triggers an exception
-        # not handled message on Python 3.8+, os._exit() does not.
-        os._exit(0)
-
-    if ex_cls is OSError:
-      if ex.errno == errno.ENOTCONN:
-        sys.exit()
-
-    raise ex
-
-if hasattr(sys,'unraisablehook'): # Python 3.8+
-    sys.unraisablehook = handle_unraisableException
-sys.excepthook = handle_Exception
-
 
 def cleanup_line(single_line):
     """Format a raw line of data from /proc and isolate the name value
@@ -612,20 +593,6 @@ def section_arc(kstats_dict):
     prt_i1('Deleted:', f_hits(arc_stats['deleted']))
     prt_i1('Mutex misses:', f_hits(arc_stats['mutex_miss']))
     prt_i1('Eviction skips:', f_hits(arc_stats['evict_skip']))
-    prt_i1('Eviction skips due to L2 writes:',
-           f_hits(arc_stats['evict_l2_skip']))
-    prt_i1('L2 cached evictions:', f_bytes(arc_stats['evict_l2_cached']))
-    prt_i1('L2 eligible evictions:', f_bytes(arc_stats['evict_l2_eligible']))
-    prt_i2('L2 eligible MFU evictions:',
-           f_perc(arc_stats['evict_l2_eligible_mfu'],
-           arc_stats['evict_l2_eligible']),
-           f_bytes(arc_stats['evict_l2_eligible_mfu']))
-    prt_i2('L2 eligible MRU evictions:',
-           f_perc(arc_stats['evict_l2_eligible_mru'],
-           arc_stats['evict_l2_eligible']),
-           f_bytes(arc_stats['evict_l2_eligible_mru']))
-    prt_i1('L2 ineligible evictions:',
-           f_bytes(arc_stats['evict_l2_ineligible']))
     print()
 
 
@@ -764,21 +731,6 @@ def section_l2arc(kstats_dict):
     prt_i2('Header size:',
            f_perc(arc_stats['l2_hdr_size'], arc_stats['l2_size']),
            f_bytes(arc_stats['l2_hdr_size']))
-    prt_i2('MFU allocated size:',
-           f_perc(arc_stats['l2_mfu_asize'], arc_stats['l2_asize']),
-           f_bytes(arc_stats['l2_mfu_asize']))
-    prt_i2('MRU allocated size:',
-           f_perc(arc_stats['l2_mru_asize'], arc_stats['l2_asize']),
-           f_bytes(arc_stats['l2_mru_asize']))
-    prt_i2('Prefetch allocated size:',
-           f_perc(arc_stats['l2_prefetch_asize'], arc_stats['l2_asize']),
-           f_bytes(arc_stats['l2_prefetch_asize']))
-    prt_i2('Data (buffer content) allocated size:',
-           f_perc(arc_stats['l2_bufc_data_asize'], arc_stats['l2_asize']),
-           f_bytes(arc_stats['l2_bufc_data_asize']))
-    prt_i2('Metadata (buffer content) allocated size:',
-           f_perc(arc_stats['l2_bufc_metadata_asize'], arc_stats['l2_asize']),
-           f_bytes(arc_stats['l2_bufc_metadata_asize']))
 
     print()
     prt_1('L2ARC breakdown:', f_hits(l2_access_total))

cmd/arcstat/arcstat.in

@@ -88,12 +88,6 @@ cols = {
     "mfug":       [4, 1000, "MFU ghost list hits per second"],
     "mrug":       [4, 1000, "MRU ghost list hits per second"],
     "eskip":      [5, 1000, "evict_skip per second"],
-    "el2skip":    [7, 1000, "evict skip, due to l2 writes, per second"],
-    "el2cach":    [7, 1024, "Size of L2 cached evictions per second"],
-    "el2el":      [5, 1024, "Size of L2 eligible evictions per second"],
-    "el2mfu":     [6, 1024, "Size of L2 eligible MFU evictions per second"],
-    "el2mru":     [6, 1024, "Size of L2 eligible MRU evictions per second"],
-    "el2inel":    [7, 1024, "Size of L2 ineligible evictions per second"],
     "mtxmis":     [6, 1000, "mutex_miss per second"],
     "dread":      [5, 1000, "Demand accesses per second"],
     "pread":      [5, 1000, "Prefetch accesses per second"],
@@ -102,16 +96,6 @@ cols = {
     "l2read":     [6, 1000, "Total L2ARC accesses per second"],
     "l2hit%":     [6, 100, "L2ARC access hit percentage"],
     "l2miss%":    [7, 100, "L2ARC access miss percentage"],
-    "l2pref":     [6, 1024, "L2ARC prefetch allocated size"],
-    "l2mfu":      [5, 1024, "L2ARC MFU allocated size"],
-    "l2mru":      [5, 1024, "L2ARC MRU allocated size"],
-    "l2data":     [6, 1024, "L2ARC data allocated size"],
-    "l2meta":     [6, 1024, "L2ARC metadata allocated size"],
-    "l2pref%":    [7, 100, "L2ARC prefetch percentage"],
-    "l2mfu%":     [6, 100, "L2ARC MFU percentage"],
-    "l2mru%":     [6, 100, "L2ARC MRU percentage"],
-    "l2data%":    [7, 100, "L2ARC data percentage"],
-    "l2meta%":    [7, 100, "L2ARC metadata percentage"],
     "l2asize":    [7, 1024, "Actual (compressed) size of the L2ARC"],
     "l2size":     [6, 1024, "Size of the L2ARC"],
     "l2bytes":    [7, 1024, "Bytes read per second from the L2ARC"],
@@ -479,12 +463,6 @@ def calculate():
     v["mrug"] = d["mru_ghost_hits"] / sint
     v["mfug"] = d["mfu_ghost_hits"] / sint
     v["eskip"] = d["evict_skip"] / sint
-    v["el2skip"] = d["evict_l2_skip"] / sint
-    v["el2cach"] = d["evict_l2_cached"] / sint
-    v["el2el"] = d["evict_l2_eligible"] / sint
-    v["el2mfu"] = d["evict_l2_eligible_mfu"] / sint
-    v["el2mru"] = d["evict_l2_eligible_mru"] / sint
-    v["el2inel"] = d["evict_l2_ineligible"] / sint
     v["mtxmis"] = d["mutex_miss"] / sint
 
     if l2exist:
@@ -498,17 +476,6 @@ def calculate():
         v["l2size"] = cur["l2_size"]
         v["l2bytes"] = d["l2_read_bytes"] / sint
 
-        v["l2pref"] = cur["l2_prefetch_asize"]
-        v["l2mfu"] = cur["l2_mfu_asize"]
-        v["l2mru"] = cur["l2_mru_asize"]
-        v["l2data"] = cur["l2_bufc_data_asize"]
-        v["l2meta"] = cur["l2_bufc_metadata_asize"]
-        v["l2pref%"] = 100 * v["l2pref"] / v["l2asize"]
-        v["l2mfu%"] = 100 * v["l2mfu"] / v["l2asize"]
-        v["l2mru%"] = 100 * v["l2mru"] / v["l2asize"]
-        v["l2data%"] = 100 * v["l2data"] / v["l2asize"]
-        v["l2meta%"] = 100 * v["l2meta"] / v["l2asize"]
-
     v["grow"] = 0 if cur["arc_no_grow"] else 1
     v["need"] = cur["arc_need_free"]
     v["free"] = cur["memory_free_bytes"]

cmd/fsck_zfs/.gitignore

@@ -1 +0,0 @@
-/fsck.zfs

cmd/fsck_zfs/Makefile.am

@@ -1,6 +1 @@
-include $(top_srcdir)/config/Substfiles.am
-include $(top_srcdir)/config/Shellcheck.am
-
 dist_sbin_SCRIPTS = fsck.zfs
-
-SUBSTFILES += $(dist_sbin_SCRIPTS)

cmd/fsck_zfs/fsck.zfs

@@ -0,0 +1,9 @@
+#!/bin/sh
+#
+# fsck.zfs: A fsck helper to accommodate distributions that expect
+# to be able to execute a fsck on all filesystem types.  Currently
+# this script does nothing but it could be extended to act as a
+# compatibility wrapper for 'zpool scrub'.
+#
+
+exit 0

cmd/fsck_zfs/fsck.zfs.in

@@ -1,44 +0,0 @@
-#!/bin/sh
-#
-# fsck.zfs: A fsck helper to accommodate distributions that expect
-# to be able to execute a fsck on all filesystem types.
-#
-# This script simply bubbles up some already-known-about errors,
-# see fsck.zfs(8)
-#
-
-if [ "$#" = "0" ]; then
-	echo "Usage: $0 [options] dataset…" >&2
-	exit 16
-fi
-
-ret=0
-for dataset in "$@"; do
-	case "$dataset" in
-		-*)
-			continue
-			;;
-		*)
-			;;
-	esac
-
-	pool="${dataset%%/*}"
-
-	case "$(@sbindir@/zpool list -Ho health "$pool")" in
-		DEGRADED)
-			ret=$(( ret | 4 ))
-			;;
-		FAULTED)
-			awk '!/^([[:space:]]*#.*)?$/ && $1 == "'"$dataset"'" && $3 == "zfs" {exit 1}' /etc/fstab || \
-				ret=$(( ret | 8 ))
-			;;
-		"")
-			# Pool not found, error printed by zpool(8)
-			ret=$(( ret | 8 ))
-			;;
-		*)
-			;;
-	esac
-done
-
-exit "$ret"

cmd/mount_zfs/mount_zfs.c

@@ -185,11 +185,10 @@ main(int argc, char **argv)
 			break;
 		case 'h':
 		case '?':
-			if (optopt)
+			(void) fprintf(stderr, gettext("Invalid option '%c'\n"),
-				(void) fprintf(stderr,
+			    optopt);
-				    gettext("Invalid option '%c'\n"), optopt);
 			(void) fprintf(stderr, gettext("Usage: mount.zfs "
-			    "[-sfnvh] [-o options] <dataset> <mountpoint>\n"));
+			    "[-sfnv] [-o options] <dataset> <mountpoint>\n"));
 			return (MOUNT_USAGE);
 		}
 	}
@@ -246,6 +245,13 @@ main(int argc, char **argv)
 		}
 	}
 
+	if (verbose)
+		(void) fprintf(stdout, gettext("mount.zfs:\n"
+		    "  dataset:    \"%s\"\n  mountpoint: \"%s\"\n"
+		    "  mountflags: 0x%lx\n  zfsflags:   0x%lx\n"
+		    "  mountopts:  \"%s\"\n  mtabopts:   \"%s\"\n"),
+		    dataset, mntpoint, mntflags, zfsflags, mntopts, mtabopt);
+
 	if (mntflags & MS_REMOUNT) {
 		nomtab = 1;
 		remount = 1;
@@ -268,10 +274,7 @@ main(int argc, char **argv)
 		return (MOUNT_USAGE);
 	}
 
-	if (!zfsutil || sloppy ||
+	zfs_adjust_mount_options(zhp, mntpoint, mntopts, mtabopt);
-	    libzfs_envvar_is_set("ZFS_MOUNT_HELPER")) {
-		zfs_adjust_mount_options(zhp, mntpoint, mntopts, mtabopt);
-	}
 
 	/* treat all snapshots as legacy mount points */
 	if (zfs_get_type(zhp) == ZFS_TYPE_SNAPSHOT)
@@ -289,11 +292,12 @@ main(int argc, char **argv)
 	if (zfs_version == 0) {
 		fprintf(stderr, gettext("unable to fetch "
 		    "ZFS version for filesystem '%s'\n"), dataset);
-		zfs_close(zhp);
-		libzfs_fini(g_zfs);
 		return (MOUNT_SYSERR);
 	}
 
+	zfs_close(zhp);
+	libzfs_fini(g_zfs);
+
 	/*
 	 * Legacy mount points may only be mounted using 'mount', never using
 	 * 'zfs mount'.  However, since 'zfs mount' actually invokes 'mount'
@@ -311,8 +315,6 @@ main(int argc, char **argv)
 		    "Use 'zfs set mountpoint=%s' or 'mount -t zfs %s %s'.\n"
 		    "See zfs(8) for more information.\n"),
 		    dataset, mntpoint, dataset, mntpoint);
-		zfs_close(zhp);
-		libzfs_fini(g_zfs);
 		return (MOUNT_USAGE);
 	}
 
@@ -323,38 +325,14 @@ main(int argc, char **argv)
 		    "Use 'zfs set mountpoint=%s' or 'zfs mount %s'.\n"
 		    "See zfs(8) for more information.\n"),
 		    dataset, "legacy", dataset);
-		zfs_close(zhp);
-		libzfs_fini(g_zfs);
 		return (MOUNT_USAGE);
 	}
 
-	if (verbose)
-		(void) fprintf(stdout, gettext("mount.zfs:\n"
-		    "  dataset:    \"%s\"\n  mountpoint: \"%s\"\n"
-		    "  mountflags: 0x%lx\n  zfsflags:   0x%lx\n"
-		    "  mountopts:  \"%s\"\n  mtabopts:   \"%s\"\n"),
-		    dataset, mntpoint, mntflags, zfsflags, mntopts, mtabopt);
-
 	if (!fake) {
-		if (zfsutil && !sloppy &&
+		error = mount(dataset, mntpoint, MNTTYPE_ZFS,
-		    !libzfs_envvar_is_set("ZFS_MOUNT_HELPER")) {
+		    mntflags, mntopts);
-			error = zfs_mount_at(zhp, mntopts, mntflags, mntpoint);
-			if (error) {
-				(void) fprintf(stderr, "zfs_mount_at() failed: "
-				    "%s", libzfs_error_description(g_zfs));
-				zfs_close(zhp);
-				libzfs_fini(g_zfs);
-				return (MOUNT_SYSERR);
-			}
-		} else {
-			error = mount(dataset, mntpoint, MNTTYPE_ZFS,
-			    mntflags, mntopts);
-		}
 	}
 
-	zfs_close(zhp);
-	libzfs_fini(g_zfs);
-
 	if (error) {
 		switch (errno) {
 		case ENOENT:
@@ -389,7 +367,7 @@ main(int argc, char **argv)
 				    "mount the filesystem again.\n"), dataset);
 				return (MOUNT_SYSERR);
 			}
-			fallthrough;
+			/* fallthru */
 #endif
 		default:
 			(void) fprintf(stderr, gettext("filesystem "

cmd/raidz_test/raidz_bench.c

@@ -31,6 +31,8 @@
 #include <sys/vdev_raidz_impl.h>
 #include <stdio.h>
 
+#include <sys/time.h>
+
 #include "raidz_test.h"
 
 #define	GEN_BENCH_MEMORY	(((uint64_t)1ULL)<<32)
@@ -81,17 +83,8 @@ run_gen_bench_impl(const char *impl)
 			/* create suitable raidz_map */
 			ncols = rto_opts.rto_dcols + fn + 1;
 			zio_bench.io_size = 1ULL << ds;
-
+			rm_bench = vdev_raidz_map_alloc(&zio_bench,
-			if (rto_opts.rto_expand) {
+			    BENCH_ASHIFT, ncols, fn+1);
-				rm_bench = vdev_raidz_map_alloc_expanded(
-				    zio_bench.io_abd,
-				    zio_bench.io_size, zio_bench.io_offset,
-				    rto_opts.rto_ashift, ncols+1, ncols,
-				    fn+1, rto_opts.rto_expand_offset);
-			} else {
-				rm_bench = vdev_raidz_map_alloc(&zio_bench,
-				    BENCH_ASHIFT, ncols, fn+1);
-			}
 
 			/* estimate iteration count */
 			iter_cnt = GEN_BENCH_MEMORY;
@@ -170,16 +163,8 @@ run_rec_bench_impl(const char *impl)
 			    (1ULL << BENCH_ASHIFT))
 				continue;
 
-			if (rto_opts.rto_expand) {
+			rm_bench = vdev_raidz_map_alloc(&zio_bench,
-				rm_bench = vdev_raidz_map_alloc_expanded(
+			    BENCH_ASHIFT, ncols, PARITY_PQR);
-				    zio_bench.io_abd,
-				    zio_bench.io_size, zio_bench.io_offset,
-				    BENCH_ASHIFT, ncols+1, ncols,
-				    PARITY_PQR, rto_opts.rto_expand_offset);
-			} else {
-				rm_bench = vdev_raidz_map_alloc(&zio_bench,
-				    BENCH_ASHIFT, ncols, PARITY_PQR);
-			}
 
 			/* estimate iteration count */
 			iter_cnt = (REC_BENCH_MEMORY);

cmd/raidz_test/raidz_test.c

@@ -77,20 +77,16 @@ static void print_opts(raidz_test_opts_t *opts, boolean_t force)
 		(void) fprintf(stdout, DBLSEP "Running with options:\n"
 		    "  (-a) zio ashift                   : %zu\n"
 		    "  (-o) zio offset                   : 1 << %zu\n"
-		    "  (-e) expanded map                 : %s\n"
-		    "  (-r) reflow offset                : %llx\n"
 		    "  (-d) number of raidz data columns : %zu\n"
 		    "  (-s) size of DATA                 : 1 << %zu\n"
 		    "  (-S) sweep parameters             : %s \n"
 		    "  (-v) verbose                      : %s \n\n",
-		    opts->rto_ashift,				/* -a */
+		    opts->rto_ashift,			/* -a */
-		    ilog2(opts->rto_offset),			/* -o */
+		    ilog2(opts->rto_offset),		/* -o */
-		    opts->rto_expand ? "yes" : "no",		/* -e */
+		    opts->rto_dcols,			/* -d */
-		    (u_longlong_t)opts->rto_expand_offset,	/* -r */
+		    ilog2(opts->rto_dsize),		/* -s */
-		    opts->rto_dcols,				/* -d */
+		    opts->rto_sweep ? "yes" : "no",	/* -S */
-		    ilog2(opts->rto_dsize),			/* -s */
+		    verbose);				/* -v */
-		    opts->rto_sweep ? "yes" : "no",		/* -S */
-		    verbose);					/* -v */
 	}
 }
 
@@ -108,8 +104,6 @@ static void usage(boolean_t requested)
 	    "\t[-S parameter sweep (default: %s)]\n"
 	    "\t[-t timeout for parameter sweep test]\n"
 	    "\t[-B benchmark all raidz implementations]\n"
-	    "\t[-e use expanded raidz map (default: %s)]\n"
-	    "\t[-r expanded raidz map reflow offset (default: %llx)]\n"
 	    "\t[-v increase verbosity (default: %zu)]\n"
 	    "\t[-h (print help)]\n"
 	    "\t[-T test the test, see if failure would be detected]\n"
@@ -120,8 +114,6 @@ static void usage(boolean_t requested)
 	    o->rto_dcols,				/* -d */
 	    ilog2(o->rto_dsize),			/* -s */
 	    rto_opts.rto_sweep ? "yes" : "no",		/* -S */
-	    rto_opts.rto_expand ? "yes" : "no",		/* -e */
-	    (u_longlong_t)o->rto_expand_offset,		/* -r */
 	    o->rto_v);					/* -d */
 
 	exit(requested ? 0 : 1);
@@ -136,7 +128,7 @@ static void process_options(int argc, char **argv)
 
 	bcopy(&rto_opts_defaults, o, sizeof (*o));
 
-	while ((opt = getopt(argc, argv, "TDBSvha:er:o:d:s:t:")) != -1) {
+	while ((opt = getopt(argc, argv, "TDBSvha:o:d:s:t:")) != -1) {
 		value = 0;
 
 		switch (opt) {
@@ -144,12 +136,6 @@ static void process_options(int argc, char **argv)
 			value = strtoull(optarg, NULL, 0);
 			o->rto_ashift = MIN(13, MAX(9, value));
 			break;
-		case 'e':
-			o->rto_expand = 1;
-			break;
-		case 'r':
-			o->rto_expand_offset = strtoull(optarg, NULL, 0);
-			break;
 		case 'o':
 			value = strtoull(optarg, NULL, 0);
 			o->rto_offset = ((1ULL << MIN(12, value)) >> 9) << 9;
@@ -193,34 +179,25 @@ static void process_options(int argc, char **argv)
 	}
 }
 
-#define	DATA_COL(rr, i) ((rr)->rr_col[rr->rr_firstdatacol + (i)].rc_abd)
+#define	DATA_COL(rm, i) ((rm)->rm_col[raidz_parity(rm) + (i)].rc_abd)
-#define	DATA_COL_SIZE(rr, i) ((rr)->rr_col[rr->rr_firstdatacol + (i)].rc_size)
+#define	DATA_COL_SIZE(rm, i) ((rm)->rm_col[raidz_parity(rm) + (i)].rc_size)
 
-#define	CODE_COL(rr, i) ((rr)->rr_col[(i)].rc_abd)
+#define	CODE_COL(rm, i) ((rm)->rm_col[(i)].rc_abd)
-#define	CODE_COL_SIZE(rr, i) ((rr)->rr_col[(i)].rc_size)
+#define	CODE_COL_SIZE(rm, i) ((rm)->rm_col[(i)].rc_size)
 
 static int
 cmp_code(raidz_test_opts_t *opts, const raidz_map_t *rm, const int parity)
 {
-	int r, i, ret = 0;
+	int i, ret = 0;
 
 	VERIFY(parity >= 1 && parity <= 3);
 
-	for (r = 0; r < rm->rm_nrows; r++) {
+	for (i = 0; i < parity; i++) {
-		raidz_row_t * const rr = rm->rm_row[r];
+		if (abd_cmp(CODE_COL(rm, i), CODE_COL(opts->rm_golden, i))
-		raidz_row_t * const rrg = opts->rm_golden->rm_row[r];
+		    != 0) {
-		for (i = 0; i < parity; i++) {
+			ret++;
-			if (CODE_COL_SIZE(rrg, i) == 0) {
+			LOG_OPT(D_DEBUG, opts,
-				VERIFY0(CODE_COL_SIZE(rr, i));
+			    "\nParity block [%d] different!\n", i);
-				continue;
-			}
-
-			if (abd_cmp(CODE_COL(rr, i),
-			    CODE_COL(rrg, i)) != 0) {
-				ret++;
-				LOG_OPT(D_DEBUG, opts,
-				    "\nParity block [%d] different!\n", i);
-			}
 		}
 	}
 	return (ret);
@@ -229,26 +206,16 @@ cmp_code(raidz_test_opts_t *opts, const raidz_map_t *rm, const int parity)
 static int
 cmp_data(raidz_test_opts_t *opts, raidz_map_t *rm)
 {
-	int r, i, dcols, ret = 0;
+	int i, ret = 0;
+	int dcols = opts->rm_golden->rm_cols - raidz_parity(opts->rm_golden);
 
-	for (r = 0; r < rm->rm_nrows; r++) {
+	for (i = 0; i < dcols; i++) {
-		raidz_row_t *rr = rm->rm_row[r];
+		if (abd_cmp(DATA_COL(opts->rm_golden, i), DATA_COL(rm, i))
-		raidz_row_t *rrg = opts->rm_golden->rm_row[r];
+		    != 0) {
-		dcols = opts->rm_golden->rm_row[0]->rr_cols -
+			ret++;
-		    raidz_parity(opts->rm_golden);
-		for (i = 0; i < dcols; i++) {
-			if (DATA_COL_SIZE(rrg, i) == 0) {
-				VERIFY0(DATA_COL_SIZE(rr, i));
-				continue;
-			}
 
-			if (abd_cmp(DATA_COL(rrg, i),
+			LOG_OPT(D_DEBUG, opts,
-			    DATA_COL(rr, i)) != 0) {
+			    "\nData block [%d] different!\n", i);
-				ret++;
-
-				LOG_OPT(D_DEBUG, opts,
-				    "\nData block [%d] different!\n", i);
-			}
 		}
 	}
 	return (ret);
@@ -269,13 +236,12 @@ init_rand(void *data, size_t size, void *private)
 static void
 corrupt_colums(raidz_map_t *rm, const int *tgts, const int cnt)
 {
-	for (int r = 0; r < rm->rm_nrows; r++) {
+	int i;
-		raidz_row_t *rr = rm->rm_row[r];
+	raidz_col_t *col;
-		for (int i = 0; i < cnt; i++) {
+
-			raidz_col_t *col = &rr->rr_col[tgts[i]];
+	for (i = 0; i < cnt; i++) {
-			abd_iterate_func(col->rc_abd, 0, col->rc_size,
+		col = &rm->rm_col[tgts[i]];
-			    init_rand, NULL);
+		abd_iterate_func(col->rc_abd, 0, col->rc_size, init_rand, NULL);
-		}
 	}
 }
 
@@ -322,22 +288,10 @@ init_raidz_golden_map(raidz_test_opts_t *opts, const int parity)
 
 	VERIFY0(vdev_raidz_impl_set("original"));
 
-	if (opts->rto_expand) {
+	opts->rm_golden = vdev_raidz_map_alloc(opts->zio_golden,
-		opts->rm_golden =
+	    opts->rto_ashift, total_ncols, parity);
-		    vdev_raidz_map_alloc_expanded(opts->zio_golden->io_abd,
+	rm_test = vdev_raidz_map_alloc(zio_test,
-		    opts->zio_golden->io_size, opts->zio_golden->io_offset,
+	    opts->rto_ashift, total_ncols, parity);
-		    opts->rto_ashift, total_ncols+1, total_ncols,
-		    parity, opts->rto_expand_offset);
-		rm_test = vdev_raidz_map_alloc_expanded(zio_test->io_abd,
-		    zio_test->io_size, zio_test->io_offset,
-		    opts->rto_ashift, total_ncols+1, total_ncols,
-		    parity, opts->rto_expand_offset);
-	} else {
-		opts->rm_golden = vdev_raidz_map_alloc(opts->zio_golden,
-		    opts->rto_ashift, total_ncols, parity);
-		rm_test = vdev_raidz_map_alloc(zio_test,
-		    opts->rto_ashift, total_ncols, parity);
-	}
 
 	VERIFY(opts->zio_golden);
 	VERIFY(opts->rm_golden);
@@ -358,187 +312,6 @@ init_raidz_golden_map(raidz_test_opts_t *opts, const int parity)
 	return (err);
 }
 
-/*
- * If reflow is not in progress, reflow_offset should be UINT64_MAX.
- * For each row, if the row is entirely before reflow_offset, it will
- * come from the new location.  Otherwise this row will come from the
- * old location.  Therefore, rows that straddle the reflow_offset will
- * come from the old location.
- *
- * NOTE: Until raidz expansion is implemented this function is only
- * needed by raidz_test.c to the multi-row raid_map_t functionality.
- */
-raidz_map_t *
-vdev_raidz_map_alloc_expanded(abd_t *abd, uint64_t size, uint64_t offset,
-    uint64_t ashift, uint64_t physical_cols, uint64_t logical_cols,
-    uint64_t nparity, uint64_t reflow_offset)
-{
-	/* The zio's size in units of the vdev's minimum sector size. */
-	uint64_t s = size >> ashift;
-	uint64_t q, r, bc, devidx, asize = 0, tot;
-
-	/*
-	 * "Quotient": The number of data sectors for this stripe on all but
-	 * the "big column" child vdevs that also contain "remainder" data.
-	 * AKA "full rows"
-	 */
-	q = s / (logical_cols - nparity);
-
-	/*
-	 * "Remainder": The number of partial stripe data sectors in this I/O.
-	 * This will add a sector to some, but not all, child vdevs.
-	 */
-	r = s - q * (logical_cols - nparity);
-
-	/* The number of "big columns" - those which contain remainder data. */
-	bc = (r == 0 ? 0 : r + nparity);
-
-	/*
-	 * The total number of data and parity sectors associated with
-	 * this I/O.
-	 */
-	tot = s + nparity * (q + (r == 0 ? 0 : 1));
-
-	/* How many rows contain data (not skip) */
-	uint64_t rows = howmany(tot, logical_cols);
-	int cols = MIN(tot, logical_cols);
-
-	raidz_map_t *rm = kmem_zalloc(offsetof(raidz_map_t, rm_row[rows]),
-	    KM_SLEEP);
-	rm->rm_nrows = rows;
-
-	for (uint64_t row = 0; row < rows; row++) {
-		raidz_row_t *rr = kmem_alloc(offsetof(raidz_row_t,
-		    rr_col[cols]), KM_SLEEP);
-		rm->rm_row[row] = rr;
-
-		/* The starting RAIDZ (parent) vdev sector of the row. */
-		uint64_t b = (offset >> ashift) + row * logical_cols;
-
-		/*
-		 * If we are in the middle of a reflow, and any part of this
-		 * row has not been copied, then use the old location of
-		 * this row.
-		 */
-		int row_phys_cols = physical_cols;
-		if (b + (logical_cols - nparity) > reflow_offset >> ashift)
-			row_phys_cols--;
-
-		/* starting child of this row */
-		uint64_t child_id = b % row_phys_cols;
-		/* The starting byte offset on each child vdev. */
-		uint64_t child_offset = (b / row_phys_cols) << ashift;
-
-		/*
-		 * We set cols to the entire width of the block, even
-		 * if this row is shorter.  This is needed because parity
-		 * generation (for Q and R) needs to know the entire width,
-		 * because it treats the short row as though it was
-		 * full-width (and the "phantom" sectors were zero-filled).
-		 *
-		 * Another approach to this would be to set cols shorter
-		 * (to just the number of columns that we might do i/o to)
-		 * and have another mechanism to tell the parity generation
-		 * about the "entire width".  Reconstruction (at least
-		 * vdev_raidz_reconstruct_general()) would also need to
-		 * know about the "entire width".
-		 */
-		rr->rr_cols = cols;
-		rr->rr_bigcols = bc;
-		rr->rr_missingdata = 0;
-		rr->rr_missingparity = 0;
-		rr->rr_firstdatacol = nparity;
-		rr->rr_abd_empty = NULL;
-		rr->rr_nempty = 0;
-
-		for (int c = 0; c < rr->rr_cols; c++, child_id++) {
-			if (child_id >= row_phys_cols) {
-				child_id -= row_phys_cols;
-				child_offset += 1ULL << ashift;
-			}
-			rr->rr_col[c].rc_devidx = child_id;
-			rr->rr_col[c].rc_offset = child_offset;
-			rr->rr_col[c].rc_orig_data = NULL;
-			rr->rr_col[c].rc_error = 0;
-			rr->rr_col[c].rc_tried = 0;
-			rr->rr_col[c].rc_skipped = 0;
-			rr->rr_col[c].rc_need_orig_restore = B_FALSE;
-
-			uint64_t dc = c - rr->rr_firstdatacol;
-			if (c < rr->rr_firstdatacol) {
-				rr->rr_col[c].rc_size = 1ULL << ashift;
-				rr->rr_col[c].rc_abd =
-				    abd_alloc_linear(rr->rr_col[c].rc_size,
-				    B_TRUE);
-			} else if (row == rows - 1 && bc != 0 && c >= bc) {
-				/*
-				 * Past the end, this for parity generation.
-				 */
-				rr->rr_col[c].rc_size = 0;
-				rr->rr_col[c].rc_abd = NULL;
-			} else {
-				/*
-				 * "data column" (col excluding parity)
-				 * Add an ASCII art diagram here
-				 */
-				uint64_t off;
-
-				if (c < bc || r == 0) {
-					off = dc * rows + row;
-				} else {
-					off = r * rows +
-					    (dc - r) * (rows - 1) + row;
-				}
-				rr->rr_col[c].rc_size = 1ULL << ashift;
-				rr->rr_col[c].rc_abd = abd_get_offset_struct(
-				    &rr->rr_col[c].rc_abdstruct,
-				    abd, off << ashift, 1 << ashift);
-			}
-
-			asize += rr->rr_col[c].rc_size;
-		}
-		/*
-		 * If all data stored spans all columns, there's a danger that
-		 * parity will always be on the same device and, since parity
-		 * isn't read during normal operation, that that device's I/O
-		 * bandwidth won't be used effectively. We therefore switch
-		 * the parity every 1MB.
-		 *
-		 * ...at least that was, ostensibly, the theory. As a practical
-		 * matter unless we juggle the parity between all devices
-		 * evenly, we won't see any benefit. Further, occasional writes
-		 * that aren't a multiple of the LCM of the number of children
-		 * and the minimum stripe width are sufficient to avoid pessimal
-		 * behavior. Unfortunately, this decision created an implicit
-		 * on-disk format requirement that we need to support for all
-		 * eternity, but only for single-parity RAID-Z.
-		 *
-		 * If we intend to skip a sector in the zeroth column for
-		 * padding we must make sure to note this swap. We will never
-		 * intend to skip the first column since at least one data and
-		 * one parity column must appear in each row.
-		 */
-		if (rr->rr_firstdatacol == 1 && rr->rr_cols > 1 &&
-		    (offset & (1ULL << 20))) {
-			ASSERT(rr->rr_cols >= 2);
-			ASSERT(rr->rr_col[0].rc_size == rr->rr_col[1].rc_size);
-			devidx = rr->rr_col[0].rc_devidx;
-			uint64_t o = rr->rr_col[0].rc_offset;
-			rr->rr_col[0].rc_devidx = rr->rr_col[1].rc_devidx;
-			rr->rr_col[0].rc_offset = rr->rr_col[1].rc_offset;
-			rr->rr_col[1].rc_devidx = devidx;
-			rr->rr_col[1].rc_offset = o;
-		}
-
-	}
-	ASSERT3U(asize, ==, tot << ashift);
-
-	/* init RAIDZ parity ops */
-	rm->rm_ops = vdev_raidz_math_get_ops();
-
-	return (rm);
-}
-
 static raidz_map_t *
 init_raidz_map(raidz_test_opts_t *opts, zio_t **zio, const int parity)
 {
@@ -557,15 +330,8 @@ init_raidz_map(raidz_test_opts_t *opts, zio_t **zio, const int parity)
 	(*zio)->io_abd = raidz_alloc(alloc_dsize);
 	init_zio_abd(*zio);
 
-	if (opts->rto_expand) {
+	rm = vdev_raidz_map_alloc(*zio, opts->rto_ashift,
-		rm = vdev_raidz_map_alloc_expanded((*zio)->io_abd,
+	    total_ncols, parity);
-		    (*zio)->io_size, (*zio)->io_offset,
-		    opts->rto_ashift, total_ncols+1, total_ncols,
-		    parity, opts->rto_expand_offset);
-	} else {
-		rm = vdev_raidz_map_alloc(*zio, opts->rto_ashift,
-		    total_ncols, parity);
-	}
 	VERIFY(rm);
 
 	/* Make sure code columns are destroyed */
@@ -654,7 +420,7 @@ run_rec_check_impl(raidz_test_opts_t *opts, raidz_map_t *rm, const int fn)
 	if (fn < RAIDZ_REC_PQ) {
 		/* can reconstruct 1 failed data disk */
 		for (x0 = 0; x0 < opts->rto_dcols; x0++) {
-			if (x0 >= rm->rm_row[0]->rr_cols - raidz_parity(rm))
+			if (x0 >= rm->rm_cols - raidz_parity(rm))
 				continue;
 
 			/* Check if should stop */
@@ -679,11 +445,10 @@ run_rec_check_impl(raidz_test_opts_t *opts, raidz_map_t *rm, const int fn)
 	} else if (fn < RAIDZ_REC_PQR) {
 		/* can reconstruct 2 failed data disk */
 		for (x0 = 0; x0 < opts->rto_dcols; x0++) {
-			if (x0 >= rm->rm_row[0]->rr_cols - raidz_parity(rm))
+			if (x0 >= rm->rm_cols - raidz_parity(rm))
 				continue;
 			for (x1 = x0 + 1; x1 < opts->rto_dcols; x1++) {
-				if (x1 >= rm->rm_row[0]->rr_cols -
+				if (x1 >= rm->rm_cols - raidz_parity(rm))
-				    raidz_parity(rm))
 					continue;
 
 				/* Check if should stop */
@@ -710,15 +475,14 @@ run_rec_check_impl(raidz_test_opts_t *opts, raidz_map_t *rm, const int fn)
 	} else {
 		/* can reconstruct 3 failed data disk */
 		for (x0 = 0; x0 < opts->rto_dcols; x0++) {
-			if (x0 >= rm->rm_row[0]->rr_cols - raidz_parity(rm))
+			if (x0 >= rm->rm_cols - raidz_parity(rm))
 				continue;
 			for (x1 = x0 + 1; x1 < opts->rto_dcols; x1++) {
-				if (x1 >= rm->rm_row[0]->rr_cols -
+				if (x1 >= rm->rm_cols - raidz_parity(rm))
-				    raidz_parity(rm))
 					continue;
 				for (x2 = x1 + 1; x2 < opts->rto_dcols; x2++) {
-					if (x2 >= rm->rm_row[0]->rr_cols -
+					if (x2 >=
-					    raidz_parity(rm))
+					    rm->rm_cols - raidz_parity(rm))
 						continue;
 
 					/* Check if should stop */
@@ -936,8 +700,6 @@ run_sweep(void)
 		opts->rto_dcols = dcols_v[d];
 		opts->rto_offset = (1 << ashift_v[a]) * rand();
 		opts->rto_dsize = size_v[s];
-		opts->rto_expand = rto_opts.rto_expand;
-		opts->rto_expand_offset = rto_opts.rto_expand_offset;
 		opts->rto_v = 0; /* be quiet */
 
 		VERIFY3P(thread_create(NULL, 0, sweep_thread, (void *) opts,
@@ -970,7 +732,6 @@ exit:
 	return (sweep_state == SWEEP_ERROR ? SWEEP_ERROR : 0);
 }
 
-
 int
 main(int argc, char **argv)
 {

cmd/raidz_test/raidz_test.h

@@ -44,15 +44,13 @@ static const char *raidz_impl_names[] = {
 
 typedef struct raidz_test_opts {
 	size_t rto_ashift;
-	uint64_t rto_offset;
+	size_t rto_offset;
 	size_t rto_dcols;
 	size_t rto_dsize;
 	size_t rto_v;
 	size_t rto_sweep;
 	size_t rto_sweep_timeout;
 	size_t rto_benchmark;
-	size_t rto_expand;
-	uint64_t rto_expand_offset;
 	size_t rto_sanity;
 	size_t rto_gdb;
 
@@ -71,8 +69,6 @@ static const raidz_test_opts_t rto_opts_defaults = {
 	.rto_v = 0,
 	.rto_sweep = 0,
 	.rto_benchmark = 0,
-	.rto_expand = 0,
-	.rto_expand_offset = -1ULL,
 	.rto_sanity = 0,
 	.rto_gdb = 0,
 	.rto_should_stop = B_FALSE
@@ -117,7 +113,4 @@ void init_zio_abd(zio_t *zio);
 
 void run_raidz_benchmark(void);
 
-struct raidz_map *vdev_raidz_map_alloc_expanded(abd_t *, uint64_t, uint64_t,
-    uint64_t, uint64_t, uint64_t, uint64_t, uint64_t);
-
 #endif /* RAIDZ_TEST_H */

cmd/vdev_id/Makefile.am

@@ -1,3 +1 @@
-include $(top_srcdir)/config/Shellcheck.am
-
 dist_udev_SCRIPTS = vdev_id

cmd/vdev_id/vdev_id

@@ -79,34 +79,6 @@
 # channel 86:00.0 1         A
 # channel 86:00.0 0         B
 
-# #
-# # Example vdev_id.conf - multipath / multijbod-daisychaining
-# #
-#
-# multipath yes
-# multijbod yes
-#
-# #       PCI_ID  HBA PORT  CHANNEL NAME
-# channel 85:00.0 1         A
-# channel 85:00.0 0         B
-# channel 86:00.0 1         A
-# channel 86:00.0 0         B
-
-# #
-# # Example vdev_id.conf - multipath / mixed
-# #
-#
-# multipath yes
-# slot mix
-#
-# #       PCI_ID  HBA PORT  CHANNEL NAME
-# channel 85:00.0 3         A
-# channel 85:00.0 2         B
-# channel 86:00.0 3         A
-# channel 86:00.0 2         B
-# channel af:00.0 0         C
-# channel af:00.0 1         C
-
 # #
 # # Example vdev_id.conf - alias
 # #
@@ -120,10 +92,9 @@ PATH=/bin:/sbin:/usr/bin:/usr/sbin
 CONFIG=/etc/zfs/vdev_id.conf
 PHYS_PER_PORT=
 DEV=
+MULTIPATH=
 TOPOLOGY=
 BAY=
-ENCL_ID=""
-UNIQ_ENCL_ID=""
 
 usage() {
 	cat << EOF
@@ -136,25 +107,22 @@ Usage: vdev_id [-h]
   -e    Create enclose device symlinks only (/dev/by-enclosure)
   -g    Storage network topology [default="$TOPOLOGY"]
   -m    Run in multipath mode
-  -j    Run in multijbod mode
   -p    number of phy's per switch port [default=$PHYS_PER_PORT]
   -h    show this summary
 EOF
-	exit 1
+	exit 0
-	# exit with error to avoid processing usage message by a udev rule
 }
 
 map_slot() {
 	LINUX_SLOT=$1
 	CHANNEL=$2
 
-	MAPPED_SLOT=$(awk -v linux_slot="$LINUX_SLOT" -v channel="$CHANNEL" \
+	MAPPED_SLOT=`awk "\\$1 == \"slot\" && \\$2 == ${LINUX_SLOT} && \
-			'$1 == "slot" && $2 == linux_slot && \
+			\\$4 ~ /^${CHANNEL}$|^$/ { print \\$3; exit }" $CONFIG`
-			($4 ~ "^"channel"$" || $4 ~ /^$/) { print $3; exit}' $CONFIG)
 	if [ -z "$MAPPED_SLOT" ] ; then
 		MAPPED_SLOT=$LINUX_SLOT
 	fi
-	printf "%d" "${MAPPED_SLOT}"
+	printf "%d" ${MAPPED_SLOT}
 }
 
 map_channel() {
@@ -164,120 +132,40 @@ map_channel() {
 
 	case $TOPOLOGY in
 		"sas_switch")
-		MAPPED_CHAN=$(awk -v port="$PORT" \
+		MAPPED_CHAN=`awk "\\$1 == \"channel\" && \\$2 == ${PORT} \
-			'$1 == "channel" && $2 == port \
+			{ print \\$3; exit }" $CONFIG`
-			{ print $3; exit }' $CONFIG)
 		;;
 		"sas_direct"|"scsi")
-		MAPPED_CHAN=$(awk -v pciID="$PCI_ID" -v port="$PORT" \
+		MAPPED_CHAN=`awk "\\$1 == \"channel\" && \
-			'$1 == "channel" && $2 == pciID && $3 == port \
+			\\$2 == \"${PCI_ID}\" && \\$3 == ${PORT} \
-			{print $4}' $CONFIG)
+			{ print \\$4; exit }" $CONFIG`
 		;;
 	esac
-	printf "%s" "${MAPPED_CHAN}"
+	printf "%s" ${MAPPED_CHAN}
-}
-
-get_encl_id() {
-	set -- $(echo $1)
-	count=$#
-
-	i=1
-	while [ $i -le $count ] ; do
-		d=$(eval echo '$'{$i})
-		id=$(cat "/sys/class/enclosure/${d}/id")
-		ENCL_ID="${ENCL_ID} $id"
-		i=$((i + 1))
-	done
-}
-
-get_uniq_encl_id() {
-	for uuid in ${ENCL_ID}; do
-		found=0
-
-		for count in ${UNIQ_ENCL_ID}; do
-			if [ $count = $uuid ]; then
-				found=1
-				break
-			fi
-		done
-
-		if [ $found -eq 0 ]; then
-			UNIQ_ENCL_ID="${UNIQ_ENCL_ID} $uuid"
-		fi
-	done
-}
-
-# map_jbod explainer: The bsg driver knows the difference between a SAS
-# expander and fanout expander. Use hostX instance along with top-level
-# (whole enclosure) expander instances in /sys/class/enclosure and
-# matching a field in an array of expanders, using the index of the
-# matched array field as the enclosure instance, thereby making jbod IDs
-# dynamic. Avoids reliance on high overhead userspace commands like
-# multipath and lsscsi and instead uses existing sysfs data.  $HOSTCHAN
-# variable derived from devpath gymnastics in sas_handler() function.
-map_jbod() {
-	DEVEXP=$(ls -l "/sys/block/$DEV/device/" | grep enclos | awk -F/ '{print $(NF-1) }')
-	DEV=$1
-
-	# Use "set --" to create index values (Arrays)
-	set -- $(ls -l /sys/class/enclosure | grep -v "^total" | awk '{print $9}')
-	# Get count of total elements
-	JBOD_COUNT=$#
-	JBOD_ITEM=$*
-
-	# Build JBODs (enclosure)  id from sys/class/enclosure/<dev>/id
-	get_encl_id "$JBOD_ITEM"
-	# Different expander instances for each paths.
-	# Filter out and keep only unique id.
-	get_uniq_encl_id
-
-	# Identify final 'mapped jbod'
-	j=0
-	for count in ${UNIQ_ENCL_ID}; do
-		i=1
-		j=$((j + 1))
-		while [ $i -le $JBOD_COUNT ] ; do
-			d=$(eval echo '$'{$i})
-			id=$(cat "/sys/class/enclosure/${d}/id")
-			if [ "$d" = "$DEVEXP" ] && [ $id = $count ] ; then
-				MAPPED_JBOD=$j
-				break
-			fi
-			i=$((i + 1))
-		done
-	done
-
-	printf "%d" "${MAPPED_JBOD}"
 }
 
 sas_handler() {
 	if [ -z "$PHYS_PER_PORT" ] ; then
-		PHYS_PER_PORT=$(awk '$1 == "phys_per_port" \
+		PHYS_PER_PORT=`awk "\\$1 == \"phys_per_port\" \
-			{print $2; exit}' $CONFIG)
+			{print \\$2; exit}" $CONFIG`
 	fi
 	PHYS_PER_PORT=${PHYS_PER_PORT:-4}
-
+	if ! echo $PHYS_PER_PORT | grep -q -E '^[0-9]+$' ; then
-	if ! echo "$PHYS_PER_PORT" | grep -q -E '^[0-9]+$' ; then
 		echo "Error: phys_per_port value $PHYS_PER_PORT is non-numeric"
 		exit 1
 	fi
 
 	if [ -z "$MULTIPATH_MODE" ] ; then
-		MULTIPATH_MODE=$(awk '$1 == "multipath" \
+		MULTIPATH_MODE=`awk "\\$1 == \"multipath\" \
-			{print $2; exit}' $CONFIG)
+			{print \\$2; exit}" $CONFIG`
-	fi
-
-	if [ -z "$MULTIJBOD_MODE" ] ; then
-		MULTIJBOD_MODE=$(awk '$1 == "multijbod" \
-			{print $2; exit}' $CONFIG)
 	fi
 
 	# Use first running component device if we're handling a dm-mpath device
 	if [ "$MULTIPATH_MODE" = "yes" ] ; then
 		# If udev didn't tell us the UUID via DM_NAME, check /dev/mapper
 		if [ -z "$DM_NAME" ] ; then
-			DM_NAME=$(ls -l --full-time /dev/mapper |
+			DM_NAME=`ls -l --full-time /dev/mapper |
-				grep "$DEV"$ | awk '{print $9}')
+				awk "/\/$DEV$/{print \\$9}"`
 		fi
 
 		# For raw disks udev exports DEVTYPE=partition when
@@ -287,50 +175,28 @@ sas_handler() {
 		# we have to append the -part suffix directly in the
 		# helper.
 		if [ "$DEVTYPE" != "partition" ] ; then
-			# Match p[number], remove the 'p' and prepend "-part"
+			PART=`echo $DM_NAME | awk -Fp '/p/{print "-part"$2}'`
-			PART=$(echo "$DM_NAME" |
-				awk 'match($0,/p[0-9]+$/) {print "-part"substr($0,RSTART+1,RLENGTH-1)}')
 		fi
 
 		# Strip off partition information.
-		DM_NAME=$(echo "$DM_NAME" | sed 's/p[0-9][0-9]*$//')
+		DM_NAME=`echo $DM_NAME | sed 's/p[0-9][0-9]*$//'`
 		if [ -z "$DM_NAME" ] ; then
 			return
 		fi
 
-		# Utilize DM device name to gather subordinate block devices
+		# Get the raw scsi device name from multipath -ll. Strip off
-		# using sysfs to avoid userspace utilities
+		# leading pipe symbols to make field numbering consistent.
-
+		DEV=`multipath -ll $DM_NAME |
-		# If our DEVNAME is something like /dev/dm-177, then we may be
+			awk '/running/{gsub("^[|]"," "); print $3 ; exit}'`
-		# able to get our DMDEV from it.
-		DMDEV=$(echo $DEVNAME | sed 's;/dev/;;g')
-		if [ ! -e /sys/block/$DMDEV/slaves/* ] ; then
-			# It's not there, try looking in /dev/mapper
-			DMDEV=$(ls -l --full-time /dev/mapper | grep $DM_NAME |
-			awk '{gsub("../", " "); print $NF}')
-		fi
-
-		# Use sysfs pointers in /sys/block/dm-X/slaves because using
-		# userspace tools creates lots of overhead and should be avoided
-		# whenever possible. Use awk to isolate lowest instance of
-		# sd device member in dm device group regardless of string
-		# length.
-		DEV=$(ls "/sys/block/$DMDEV/slaves" | awk '
-			{ len=sprintf ("%20s",length($0)); gsub(/ /,0,str); a[NR]=len "_" $0; }
-			END {
-				asort(a)
-				print substr(a[1],22)
-			}')
-
 		if [ -z "$DEV" ] ; then
 			return
 		fi
 	fi
 
-	if echo "$DEV" | grep -q ^/devices/ ; then
+	if echo $DEV | grep -q ^/devices/ ; then
 		sys_path=$DEV
 	else
-		sys_path=$(udevadm info -q path -p "/sys/block/$DEV" 2>/dev/null)
+		sys_path=`udevadm info -q path -p /sys/block/$DEV 2>/dev/null`
 	fi
 
 	# Use positional parameters as an ad-hoc array
@@ -340,104 +206,84 @@ sas_handler() {
 
 	# Get path up to /sys/.../hostX
 	i=1
-
+	while [ $i -le $num_dirs ] ; do
-	while [ $i -le "$num_dirs" ] ; do
+		d=$(eval echo \${$i})
-		d=$(eval echo '$'{$i})
 		scsi_host_dir="$scsi_host_dir/$d"
-		echo "$d" | grep -q -E '^host[0-9]+$' && break
+		echo $d | grep -q -E '^host[0-9]+$' && break
-		i=$((i + 1))
+		i=$(($i + 1))
 	done
 
-	# Lets grab the SAS host channel number and save it for JBOD sorting later
+	if [ $i = $num_dirs ] ; then
-	HOSTCHAN=$(echo "$d" | awk -F/ '{ gsub("host","",$NF); print $NF}')
-
-	if [ $i = "$num_dirs" ] ; then
 		return
 	fi
 
-	PCI_ID=$(eval echo '$'{$((i -1))} | awk -F: '{print $2":"$3}')
+	PCI_ID=$(eval echo \${$(($i -1))} | awk -F: '{print $2":"$3}')
 
 	# In sas_switch mode, the directory four levels beneath
 	# /sys/.../hostX contains symlinks to phy devices that reveal
 	# the switch port number.  In sas_direct mode, the phy links one
 	# directory down reveal the HBA port.
 	port_dir=$scsi_host_dir
-
 	case $TOPOLOGY in
-		"sas_switch") j=$((i + 4)) ;;
+		"sas_switch") j=$(($i + 4)) ;;
-		"sas_direct") j=$((i + 1)) ;;
+		"sas_direct") j=$(($i + 1)) ;;
 	esac
 
-	i=$((i + 1))
+	i=$(($i + 1))
-
 	while [ $i -le $j ] ; do
-		port_dir="$port_dir/$(eval echo '$'{$i})"
+		port_dir="$port_dir/$(eval echo \${$i})"
-		i=$((i + 1))
+		i=$(($i + 1))
 	done
 
-	PHY=$(ls -vd "$port_dir"/phy* 2>/dev/null | head -1 | awk -F: '{print $NF}')
+	PHY=`ls -d $port_dir/phy* 2>/dev/null | head -1 | awk -F: '{print $NF}'`
 	if [ -z "$PHY" ] ; then
 		PHY=0
 	fi
-	PORT=$((PHY / PHYS_PER_PORT))
+	PORT=$(( $PHY / $PHYS_PER_PORT ))
 
 	# Look in /sys/.../sas_device/end_device-X for the bay_identifier
 	# attribute.
 	end_device_dir=$port_dir
-
+	while [ $i -lt $num_dirs ] ; do
-	while [ $i -lt "$num_dirs" ] ; do
+		d=$(eval echo \${$i})
-		d=$(eval echo '$'{$i})
 		end_device_dir="$end_device_dir/$d"
-		if echo "$d" | grep -q '^end_device' ; then
+		if echo $d | grep -q '^end_device' ; then
 			end_device_dir="$end_device_dir/sas_device/$d"
 			break
 		fi
-		i=$((i + 1))
+		i=$(($i + 1))
 	done
 
-	# Add 'mix' slot type for environments where dm-multipath devices
-	# include end-devices connected via SAS expanders or direct connection
-	# to SAS HBA. A mixed connectivity environment such as pool devices
-	# contained in a SAS JBOD and spare drives or log devices directly
-	# connected in a server backplane without expanders in the I/O path.
 	SLOT=
-
 	case $BAY in
 	"bay")
-		SLOT=$(cat "$end_device_dir/bay_identifier" 2>/dev/null)
+		SLOT=`cat $end_device_dir/bay_identifier 2>/dev/null`
-		;;
-	"mix")
-		if [ $(cat "$end_device_dir/bay_identifier" 2>/dev/null) ] ; then
-			SLOT=$(cat "$end_device_dir/bay_identifier" 2>/dev/null)
-		else
-			SLOT=$(cat "$end_device_dir/phy_identifier" 2>/dev/null)
-		fi
 		;;
 	"phy")
-		SLOT=$(cat "$end_device_dir/phy_identifier" 2>/dev/null)
+		SLOT=`cat $end_device_dir/phy_identifier 2>/dev/null`
 		;;
 	"port")
-		d=$(eval echo '$'{$i})
+		d=$(eval echo \${$i})
-		SLOT=$(echo "$d" | sed -e 's/^.*://')
+		SLOT=`echo $d | sed -e 's/^.*://'`
 		;;
 	"id")
-		i=$((i + 1))
+		i=$(($i + 1))
-		d=$(eval echo '$'{$i})
+		d=$(eval echo \${$i})
-		SLOT=$(echo "$d" | sed -e 's/^.*://')
+		SLOT=`echo $d | sed -e 's/^.*://'`
 		;;
 	"lun")
-		i=$((i + 2))
+		i=$(($i + 2))
-		d=$(eval echo '$'{$i})
+		d=$(eval echo \${$i})
-		SLOT=$(echo "$d" | sed -e 's/^.*://')
+		SLOT=`echo $d | sed -e 's/^.*://'`
 		;;
 	"ses")
 		# look for this SAS path in all SCSI Enclosure Services
 		# (SES) enclosures
-		sas_address=$(cat "$end_device_dir/sas_address" 2>/dev/null)
+		sas_address=`cat $end_device_dir/sas_address 2>/dev/null`
-		enclosures=$(lsscsi -g | \
+		enclosures=`lsscsi -g | \
-			sed -n -e '/enclosu/s/^.* \([^ ][^ ]*\) *$/\1/p')
+			sed -n -e '/enclosu/s/^.* \([^ ][^ ]*\) *$/\1/p'`
 		for enclosure in $enclosures; do
-			set -- $(sg_ses -p aes "$enclosure" | \
+			set -- $(sg_ses -p aes $enclosure | \
 				awk "/device slot number:/{slot=\$12} \
 					/SAS address: $sas_address/\
 					{print slot}")
@@ -452,55 +298,42 @@ sas_handler() {
 		return
 	fi
 
-	if [ "$MULTIJBOD_MODE" = "yes" ] ; then
+	CHAN=`map_channel $PCI_ID $PORT`
-		CHAN=$(map_channel "$PCI_ID" "$PORT")
+	SLOT=`map_slot $SLOT $CHAN`
-		SLOT=$(map_slot "$SLOT" "$CHAN")
+	if [ -z "$CHAN" ] ; then
-		JBOD=$(map_jbod "$DEV")
+		return
-
-		if [ -z "$CHAN" ] ; then
-			return
-		fi
-		echo "${CHAN}"-"${JBOD}"-"${SLOT}${PART}"
-	else
-		CHAN=$(map_channel "$PCI_ID" "$PORT")
-		SLOT=$(map_slot "$SLOT" "$CHAN")
-
-		if [ -z "$CHAN" ] ; then
-			return
-		fi
-		echo "${CHAN}${SLOT}${PART}"
 	fi
+	echo ${CHAN}${SLOT}${PART}
 }
 
 scsi_handler() {
 	if [ -z "$FIRST_BAY_NUMBER" ] ; then
-		FIRST_BAY_NUMBER=$(awk '$1 == "first_bay_number" \
+		FIRST_BAY_NUMBER=`awk "\\$1 == \"first_bay_number\" \
-			{print $2; exit}' $CONFIG)
+			{print \\$2; exit}" $CONFIG`
 	fi
 	FIRST_BAY_NUMBER=${FIRST_BAY_NUMBER:-0}
 
 	if [ -z "$PHYS_PER_PORT" ] ; then
-		PHYS_PER_PORT=$(awk '$1 == "phys_per_port" \
+		PHYS_PER_PORT=`awk "\\$1 == \"phys_per_port\" \
-			{print $2; exit}' $CONFIG)
+			{print \\$2; exit}" $CONFIG`
 	fi
 	PHYS_PER_PORT=${PHYS_PER_PORT:-4}
-
+	if ! echo $PHYS_PER_PORT | grep -q -E '^[0-9]+$' ; then
-	if ! echo "$PHYS_PER_PORT" | grep -q -E '^[0-9]+$' ; then
 		echo "Error: phys_per_port value $PHYS_PER_PORT is non-numeric"
 		exit 1
 	fi
 
 	if [ -z "$MULTIPATH_MODE" ] ; then
-		MULTIPATH_MODE=$(awk '$1 == "multipath" \
+		MULTIPATH_MODE=`awk "\\$1 == \"multipath\" \
-			{print $2; exit}' $CONFIG)
+			{print \\$2; exit}" $CONFIG`
 	fi
 
 	# Use first running component device if we're handling a dm-mpath device
 	if [ "$MULTIPATH_MODE" = "yes" ] ; then
 		# If udev didn't tell us the UUID via DM_NAME, check /dev/mapper
 		if [ -z "$DM_NAME" ] ; then
-			DM_NAME=$(ls -l --full-time /dev/mapper |
+			DM_NAME=`ls -l --full-time /dev/mapper |
-				grep "$DEV"$ | awk '{print $9}')
+				awk "/\/$DEV$/{print \\$9}"`
 		fi
 
 		# For raw disks udev exports DEVTYPE=partition when
@@ -510,30 +343,28 @@ scsi_handler() {
 		# we have to append the -part suffix directly in the
 		# helper.
 		if [ "$DEVTYPE" != "partition" ] ; then
-			# Match p[number], remove the 'p' and prepend "-part"
+			PART=`echo $DM_NAME | awk -Fp '/p/{print "-part"$2}'`
-			PART=$(echo "$DM_NAME" |
-			    awk 'match($0,/p[0-9]+$/) {print "-part"substr($0,RSTART+1,RLENGTH-1)}')
 		fi
 
 		# Strip off partition information.
-		DM_NAME=$(echo "$DM_NAME" | sed 's/p[0-9][0-9]*$//')
+		DM_NAME=`echo $DM_NAME | sed 's/p[0-9][0-9]*$//'`
 		if [ -z "$DM_NAME" ] ; then
 			return
 		fi
 
 		# Get the raw scsi device name from multipath -ll. Strip off
 		# leading pipe symbols to make field numbering consistent.
-		DEV=$(multipath -ll "$DM_NAME" |
+		DEV=`multipath -ll $DM_NAME |
-			awk '/running/{gsub("^[|]"," "); print $3 ; exit}')
+			awk '/running/{gsub("^[|]"," "); print $3 ; exit}'`
 		if [ -z "$DEV" ] ; then
 			return
 		fi
 	fi
 
-	if echo "$DEV" | grep -q ^/devices/ ; then
+	if echo $DEV | grep -q ^/devices/ ; then
 		sys_path=$DEV
 	else
-		sys_path=$(udevadm info -q path -p "/sys/block/$DEV" 2>/dev/null)
+		sys_path=`udevadm info -q path -p /sys/block/$DEV 2>/dev/null`
 	fi
 
 	# expect sys_path like this, for example:
@@ -546,47 +377,44 @@ scsi_handler() {
 
 	# Get path up to /sys/.../hostX
 	i=1
-
+	while [ $i -le $num_dirs ] ; do
-	while [ $i -le "$num_dirs" ] ; do
+		d=$(eval echo \${$i})
-		d=$(eval echo '$'{$i})
 		scsi_host_dir="$scsi_host_dir/$d"
-
+		echo $d | grep -q -E '^host[0-9]+$' && break
-		echo "$d" | grep -q -E '^host[0-9]+$' && break
+		i=$(($i + 1))
-		i=$((i + 1))
 	done
 
-	if [ $i = "$num_dirs" ] ; then
+	if [ $i = $num_dirs ] ; then
 		return
 	fi
 
-	PCI_ID=$(eval echo '$'{$((i -1))} | awk -F: '{print $2":"$3}')
+	PCI_ID=$(eval echo \${$(($i -1))} | awk -F: '{print $2":"$3}')
 
 	# In scsi mode, the directory two levels beneath
 	# /sys/.../hostX reveals the port and slot.
 	port_dir=$scsi_host_dir
-	j=$((i + 2))
+	j=$(($i + 2))
 
-	i=$((i + 1))
+	i=$(($i + 1))
 	while [ $i -le $j ] ; do
-		port_dir="$port_dir/$(eval echo '$'{$i})"
+		port_dir="$port_dir/$(eval echo \${$i})"
-		i=$((i + 1))
+		i=$(($i + 1))
 	done
 
-	set -- $(echo "$port_dir" | sed -e 's/^.*:\([^:]*\):\([^:]*\)$/\1 \2/')
+	set -- $(echo $port_dir | sed -e 's/^.*:\([^:]*\):\([^:]*\)$/\1 \2/')
 	PORT=$1
-	SLOT=$(($2 + FIRST_BAY_NUMBER))
+	SLOT=$(($2 + $FIRST_BAY_NUMBER))
 
 	if [ -z "$SLOT" ] ; then
 		return
 	fi
 
-	CHAN=$(map_channel "$PCI_ID" "$PORT")
+	CHAN=`map_channel $PCI_ID $PORT`
-	SLOT=$(map_slot "$SLOT" "$CHAN")
+	SLOT=`map_slot $SLOT $CHAN`
-
 	if [ -z "$CHAN" ] ; then
 		return
 	fi
-	echo "${CHAN}${SLOT}${PART}"
+	echo ${CHAN}${SLOT}${PART}
 }
 
 # Figure out the name for the enclosure symlink
@@ -596,10 +424,8 @@ enclosure_handler () {
 	# DEVPATH=/sys/devices/pci0000:00/0000:00:03.0/0000:05:00.0/host0/subsystem/devices/0:0:0:0/scsi_generic/sg0
 
 	# Get the enclosure ID ("0:0:0:0")
-	ENC="${DEVPATH%/*}"
+	ENC=$(basename $(readlink -m "/sys/$DEVPATH/../.."))
-	ENC="${ENC%/*}"
+	if [ ! -d /sys/class/enclosure/$ENC ] ; then
-	ENC="${ENC##*/}"
-	if [ ! -d "/sys/class/enclosure/$ENC" ] ; then
 		# Not an enclosure, bail out
 		return
 	fi
@@ -607,26 +433,25 @@ enclosure_handler () {
 	# Get the long sysfs device path to our enclosure. Looks like:
 	# /devices/pci0000:00/0000:00:03.0/0000:05:00.0/host0/port-0:0/ ... /enclosure/0:0:0:0
 
-	ENC_DEVICE=$(readlink "/sys/class/enclosure/$ENC")
+	ENC_DEVICE=$(readlink /sys/class/enclosure/$ENC)
 
 	# Grab the full path to the hosts port dir:
 	# /devices/pci0000:00/0000:00:03.0/0000:05:00.0/host0/port-0:0
-	PORT_DIR=$(echo "$ENC_DEVICE" | grep -Eo '.+host[0-9]+/port-[0-9]+:[0-9]+')
+	PORT_DIR=$(echo $ENC_DEVICE | grep -Eo '.+host[0-9]+/port-[0-9]+:[0-9]+')
 
 	# Get the port number
-	PORT_ID=$(echo "$PORT_DIR" | grep -Eo "[0-9]+$")
+	PORT_ID=$(echo $PORT_DIR | grep -Eo "[0-9]+$")
 
 	# The PCI directory is two directories up from the port directory
 	# /sys/devices/pci0000:00/0000:00:03.0/0000:05:00.0
-	PCI_ID_LONG="$(readlink -m "/sys/$PORT_DIR/../..")"
+	PCI_ID_LONG=$(basename $(readlink -m "/sys/$PORT_DIR/../.."))
-	PCI_ID_LONG="${PCI_ID_LONG##*/}"
 
 	# Strip down the PCI address from 0000:05:00.0 to 05:00.0
-	PCI_ID="${PCI_ID_LONG#[0-9]*:}"
+	PCI_ID=$(echo "$PCI_ID_LONG" | sed -r 's/^[0-9]+://g')
 
 	# Name our device according to vdev_id.conf (like "L0" or "U1").
 	NAME=$(awk "/channel/{if (\$1 == \"channel\" && \$2 == \"$PCI_ID\" && \
-		\$3 == \"$PORT_ID\") {print \$4\$3}}" $CONFIG)
+		\$3 == \"$PORT_ID\") {print \$4int(count[\$4])}; count[\$4]++}" $CONFIG)
 
 	echo "${NAME}"
 }
@@ -662,11 +487,9 @@ alias_handler () {
 	#          ambiguity seems unavoidable, so devices using this facility
 	#          must not use such names.
 	DM_PART=
-	if echo "$DM_NAME" | grep -q -E 'p[0-9][0-9]*$' ; then
+	if echo $DM_NAME | grep -q -E 'p[0-9][0-9]*$' ; then
 		if [ "$DEVTYPE" != "partition" ] ; then
-			# Match p[number], remove the 'p' and prepend "-part"
+			DM_PART=`echo $DM_NAME | awk -Fp '/p/{print "-part"$2}'`
-			DM_PART=$(echo "$DM_NAME" |
-			    awk 'match($0,/p[0-9]+$/) {print "-part"substr($0,RSTART+1,RLENGTH-1)}')
 		fi
 	fi
 
@@ -674,25 +497,21 @@ alias_handler () {
 	for link in $DEVLINKS ; do
 		# Remove partition information to match key of top-level device.
 		if [ -n "$DM_PART" ] ; then
-			link=$(echo "$link" | sed 's/p[0-9][0-9]*$//')
+			link=`echo $link | sed 's/p[0-9][0-9]*$//'`
 		fi
 		# Check both the fully qualified and the base name of link.
-		for l in $link ${link##*/} ; do
+		for l in $link `basename $link` ; do
-			if [ ! -z "$l" ]; then
+			alias=`awk "\\$1 == \"alias\" && \\$3 == \"${l}\" \
-				alias=$(awk -v var="$l" '($1 == "alias") && \
+					{ print \\$2; exit }" $CONFIG`
-					($3 == var) \
+			if [ -n "$alias" ] ; then
-					{ print $2; exit }' $CONFIG)
+				echo ${alias}${DM_PART}
-				if [ -n "$alias" ] ; then
+				return
-					echo "${alias}${DM_PART}"
-					return
-				fi
 			fi
 		done
 	done
 }
 
-# main
+while getopts 'c:d:eg:mp:h' OPTION; do
-while getopts 'c:d:eg:jmp:h' OPTION; do
 	case ${OPTION} in
 	c)
 		CONFIG=${OPTARG}
@@ -705,9 +524,7 @@ while getopts 'c:d:eg:jmp:h' OPTION; do
 	# create the enclosure device symlinks only.  We also need
 	# "enclosure_symlinks yes" set in vdev_id.config to actually create the
 	# symlink.
-	ENCLOSURE_MODE=$(awk '{if ($1 == "enclosure_symlinks") \
+	ENCLOSURE_MODE=$(awk '{if ($1 == "enclosure_symlinks") print $2}' $CONFIG)
-		print $2}' "$CONFIG")
-
 	if [ "$ENCLOSURE_MODE" != "yes" ] ; then
 		exit 0
 	fi
@@ -718,9 +535,6 @@ while getopts 'c:d:eg:jmp:h' OPTION; do
 	p)
 		PHYS_PER_PORT=${OPTARG}
 		;;
-	j)
-		MULTIJBOD_MODE=yes
-		;;
 	m)
 		MULTIPATH_MODE=yes
 		;;
@@ -730,9 +544,9 @@ while getopts 'c:d:eg:jmp:h' OPTION; do
 	esac
 done
 
-if [ ! -r "$CONFIG" ] ; then
+if [ ! -r $CONFIG ] ; then
 	echo "Error: Config file \"$CONFIG\" not found"
-	exit 1
+	exit 0
 fi
 
 if [ -z "$DEV" ] && [ -z "$ENCLOSURE_MODE" ] ; then
@@ -741,11 +555,11 @@ if [ -z "$DEV" ] && [ -z "$ENCLOSURE_MODE" ] ; then
 fi
 
 if [ -z "$TOPOLOGY" ] ; then
-	TOPOLOGY=$(awk '($1 == "topology") {print $2; exit}' "$CONFIG")
+	TOPOLOGY=`awk "\\$1 == \"topology\" {print \\$2; exit}" $CONFIG`
 fi
 
 if [ -z "$BAY" ] ; then
-	BAY=$(awk '($1 == "slot") {print $2; exit}' "$CONFIG")
+	BAY=`awk "\\$1 == \"slot\" {print \\$2; exit}" $CONFIG`
 fi
 
 TOPOLOGY=${TOPOLOGY:-sas_direct}
@@ -758,7 +572,7 @@ if [ "$ENCLOSURE_MODE" = "yes" ] && [ "$TOPOLOGY" = "sas_direct" ] ; then
 	fi
 
 	# Just create the symlinks to the enclosure devices and then exit.
-	ENCLOSURE_PREFIX=$(awk '/enclosure_symlinks_prefix/{print $2}' "$CONFIG")
+	ENCLOSURE_PREFIX=$(awk '/enclosure_symlinks_prefix/{print $2}' $CONFIG)
 	if [ -z "$ENCLOSURE_PREFIX" ] ; then
 		ENCLOSURE_PREFIX="enc"
 	fi
@@ -768,16 +582,16 @@ if [ "$ENCLOSURE_MODE" = "yes" ] && [ "$TOPOLOGY" = "sas_direct" ] ; then
 fi
 
 # First check if an alias was defined for this device.
-ID_VDEV=$(alias_handler)
+ID_VDEV=`alias_handler`
 
 if [ -z "$ID_VDEV" ] ; then
 	BAY=${BAY:-bay}
 	case $TOPOLOGY in
 		sas_direct|sas_switch)
-			ID_VDEV=$(sas_handler)
+			ID_VDEV=`sas_handler`
 			;;
 		scsi)
-			ID_VDEV=$(scsi_handler)
+			ID_VDEV=`scsi_handler`
 			;;
 		*)
 			echo "Error: unknown topology $TOPOLOGY"

cmd/zdb/zdb.c

@@ -31,7 +31,6 @@
  *
  * [1] Portions of this software were developed by Allan Jude
  *     under sponsorship from the FreeBSD Foundation.
- * Copyright (c) 2021 Allan Jude
  */
 
 #include <stdio.h>
@@ -110,7 +109,6 @@ extern int zfs_recover;
 extern unsigned long zfs_arc_meta_min, zfs_arc_meta_limit;
 extern int zfs_vdev_async_read_max_active;
 extern boolean_t spa_load_verify_dryrun;
-extern boolean_t spa_mode_readable_spacemaps;
 extern int zfs_reconstruct_indirect_combinations_max;
 extern int zfs_btree_verify_intensity;
 
@@ -784,14 +782,13 @@ usage(void)
 	    "\t%s -m [-AFLPX] [-e [-V] [-p <path> ...]] [-t <txg>] "
 	    "[-U <cache>]\n\t\t<poolname> [<vdev> [<metaslab> ...]]\n"
 	    "\t%s -O <dataset> <path>\n"
-	    "\t%s -r <dataset> <path> <destination>\n"
 	    "\t%s -R [-A] [-e [-V] [-p <path> ...]] [-U <cache>]\n"
 	    "\t\t<poolname> <vdev>:<offset>:<size>[:<flags>]\n"
 	    "\t%s -E [-A] word0:word1:...:word15\n"
 	    "\t%s -S [-AP] [-e [-V] [-p <path> ...]] [-U <cache>] "
 	    "<poolname>\n\n",
 	    cmdname, cmdname, cmdname, cmdname, cmdname, cmdname, cmdname,
-	    cmdname, cmdname, cmdname, cmdname);
+	    cmdname, cmdname, cmdname);
 
 	(void) fprintf(stderr, "    Dataset name must include at least one "
 	    "separator character '/' or '@'\n");
@@ -830,7 +827,6 @@ usage(void)
 	(void) fprintf(stderr, "        -m metaslabs\n");
 	(void) fprintf(stderr, "        -M metaslab groups\n");
 	(void) fprintf(stderr, "        -O perform object lookups by path\n");
-	(void) fprintf(stderr, "        -r copy an object by path to file\n");
 	(void) fprintf(stderr, "        -R read and display block from a "
 	    "device\n");
 	(void) fprintf(stderr, "        -s report stats on zdb's I/O\n");
@@ -1673,11 +1669,7 @@ dump_metaslab(metaslab_t *msp)
 		    SPACE_MAP_HISTOGRAM_SIZE, sm->sm_shift);
 	}
 
-	if (vd->vdev_ops == &vdev_draid_ops)
+	ASSERT(msp->ms_size == (1ULL << vd->vdev_ms_shift));
-		ASSERT3U(msp->ms_size, <=, 1ULL << vd->vdev_ms_shift);
-	else
-		ASSERT3U(msp->ms_size, ==, 1ULL << vd->vdev_ms_shift);
-
 	dump_spacemap(spa->spa_meta_objset, msp->ms_sm);
 
 	if (spa_feature_is_active(spa, SPA_FEATURE_LOG_SPACEMAP)) {
@@ -2219,8 +2211,7 @@ snprintf_zstd_header(spa_t *spa, char *blkbuf, size_t buflen,
 		(void) snprintf(blkbuf + strlen(blkbuf),
 		    buflen - strlen(blkbuf),
 		    " ZSTD:size=%u:version=%u:level=%u:EMBEDDED",
-		    zstd_hdr.c_len, zfs_get_hdrversion(&zstd_hdr),
+		    zstd_hdr.c_len, zstd_hdr.version, zstd_hdr.level);
-		    zfs_get_hdrlevel(&zstd_hdr));
 		return;
 	}
 
@@ -2244,8 +2235,7 @@ snprintf_zstd_header(spa_t *spa, char *blkbuf, size_t buflen,
 	(void) snprintf(blkbuf + strlen(blkbuf),
 	    buflen - strlen(blkbuf),
 	    " ZSTD:size=%u:version=%u:level=%u:NORMAL",
-	    zstd_hdr.c_len, zfs_get_hdrversion(&zstd_hdr),
+	    zstd_hdr.c_len, zstd_hdr.version, zstd_hdr.level);
-	    zfs_get_hdrlevel(&zstd_hdr));
 
 	abd_return_buf_copy(pabd, buf, BP_GET_LSIZE(bp));
 }
@@ -3125,18 +3115,13 @@ dump_znode_symlink(sa_handle_t *hdl)
 {
 	int sa_symlink_size = 0;
 	char linktarget[MAXPATHLEN];
+	linktarget[0] = '\0';
 	int error;
 
 	error = sa_size(hdl, sa_attr_table[ZPL_SYMLINK], &sa_symlink_size);
 	if (error || sa_symlink_size == 0) {
 		return;
 	}
-	if (sa_symlink_size >= sizeof (linktarget)) {
-		(void) printf("symlink size %d is too large\n",
-		    sa_symlink_size);
-		return;
-	}
-	linktarget[sa_symlink_size] = '\0';
 	if (sa_lookup(hdl, sa_attr_table[ZPL_SYMLINK],
 	    &linktarget, sa_symlink_size) == 0)
 		(void) printf("\ttarget	%s\n", linktarget);
@@ -4102,7 +4087,7 @@ cksum_record_compare(const void *x1, const void *x2)
 	const cksum_record_t *l = (cksum_record_t *)x1;
 	const cksum_record_t *r = (cksum_record_t *)x2;
 	int arraysize = ARRAY_SIZE(l->cksum.zc_word);
-	int difference = 0;
+	int difference;
 
 	for (int i = 0; i < arraysize; i++) {
 		difference = TREE_CMP(l->cksum.zc_word[i], r->cksum.zc_word[i]);
@@ -4244,8 +4229,6 @@ dump_l2arc_log_entries(uint64_t log_entries,
 		    (u_longlong_t)L2BLK_GET_PREFETCH((&le[j])->le_prop));
 		(void) printf("|\t\t\t\taddress: %llu\n",
 		    (u_longlong_t)le[j].le_daddr);
-		(void) printf("|\t\t\t\tARC state: %llu\n",
-		    (u_longlong_t)L2BLK_GET_STATE((&le[j])->le_prop));
 		(void) printf("|\n");
 	}
 	(void) printf("\n");
@@ -4528,7 +4511,7 @@ static char curpath[PATH_MAX];
  * for the last one.
  */
 static int
-dump_path_impl(objset_t *os, uint64_t obj, char *name, uint64_t *retobj)
+dump_path_impl(objset_t *os, uint64_t obj, char *name)
 {
 	int err;
 	boolean_t header = B_TRUE;
@@ -4578,15 +4561,10 @@ dump_path_impl(objset_t *os, uint64_t obj, char *name, uint64_t *retobj)
 	switch (doi.doi_type) {
 	case DMU_OT_DIRECTORY_CONTENTS:
 		if (s != NULL && *(s + 1) != '\0')
-			return (dump_path_impl(os, child_obj, s + 1, retobj));
+			return (dump_path_impl(os, child_obj, s + 1));
-		fallthrough;
+		/*FALLTHROUGH*/
 	case DMU_OT_PLAIN_FILE_CONTENTS:
-		if (retobj != NULL) {
+		dump_object(os, child_obj, dump_opt['v'], &header, NULL, 0);
-			*retobj = child_obj;
-		} else {
-			dump_object(os, child_obj, dump_opt['v'], &header,
-			    NULL, 0);
-		}
 		return (0);
 	default:
 		(void) fprintf(stderr, "object %llu has non-file/directory "
@@ -4601,7 +4579,7 @@ dump_path_impl(objset_t *os, uint64_t obj, char *name, uint64_t *retobj)
  * Dump the blocks for the object specified by path inside the dataset.
  */
 static int
-dump_path(char *ds, char *path, uint64_t *retobj)
+dump_path(char *ds, char *path)
 {
 	int err;
 	objset_t *os;
@@ -4621,89 +4599,12 @@ dump_path(char *ds, char *path, uint64_t *retobj)
 
 	(void) snprintf(curpath, sizeof (curpath), "dataset=%s path=/", ds);
 
-	err = dump_path_impl(os, root_obj, path, retobj);
+	err = dump_path_impl(os, root_obj, path);
 
 	close_objset(os, FTAG);
 	return (err);
 }
 
-static int
-zdb_copy_object(objset_t *os, uint64_t srcobj, char *destfile)
-{
-	int err = 0;
-	uint64_t size, readsize, oursize, offset;
-	ssize_t writesize;
-	sa_handle_t *hdl;
-
-	(void) printf("Copying object %" PRIu64 " to file %s\n", srcobj,
-	    destfile);
-
-	VERIFY3P(os, ==, sa_os);
-	if ((err = sa_handle_get(os, srcobj, NULL, SA_HDL_PRIVATE, &hdl))) {
-		(void) printf("Failed to get handle for SA znode\n");
-		return (err);
-	}
-	if ((err = sa_lookup(hdl, sa_attr_table[ZPL_SIZE], &size, 8))) {
-		(void) sa_handle_destroy(hdl);
-		return (err);
-	}
-	(void) sa_handle_destroy(hdl);
-
-	(void) printf("Object %" PRIu64 " is %" PRIu64 " bytes\n", srcobj,
-	    size);
-	if (size == 0) {
-		return (EINVAL);
-	}
-
-	int fd = open(destfile, O_WRONLY | O_CREAT | O_TRUNC, 0644);
-	/*
-	 * We cap the size at 1 mebibyte here to prevent
-	 * allocation failures and nigh-infinite printing if the
-	 * object is extremely large.
-	 */
-	oursize = MIN(size, 1 << 20);
-	offset = 0;
-	char *buf = kmem_alloc(oursize, KM_NOSLEEP);
-	if (buf == NULL) {
-		return (ENOMEM);
-	}
-
-	while (offset < size) {
-		readsize = MIN(size - offset, 1 << 20);
-		err = dmu_read(os, srcobj, offset, readsize, buf, 0);
-		if (err != 0) {
-			(void) printf("got error %u from dmu_read\n", err);
-			kmem_free(buf, oursize);
-			return (err);
-		}
-		if (dump_opt['v'] > 3) {
-			(void) printf("Read offset=%" PRIu64 " size=%" PRIu64
-			    " error=%d\n", offset, readsize, err);
-		}
-
-		writesize = write(fd, buf, readsize);
-		if (writesize < 0) {
-			err = errno;
-			break;
-		} else if (writesize != readsize) {
-			/* Incomplete write */
-			(void) fprintf(stderr, "Short write, only wrote %llu of"
-			    " %" PRIu64 " bytes, exiting...\n",
-			    (u_longlong_t)writesize, readsize);
-			break;
-		}
-
-		offset += readsize;
-	}
-
-	(void) close(fd);
-
-	if (buf != NULL)
-		kmem_free(buf, oursize);
-
-	return (err);
-}
-
 static int
 dump_label(const char *dev)
 {
@@ -5327,6 +5228,8 @@ zdb_blkptr_done(zio_t *zio)
 	zdb_cb_t *zcb = zio->io_private;
 	zbookmark_phys_t *zb = &zio->io_bookmark;
 
+	abd_free(zio->io_abd);
+
 	mutex_enter(&spa->spa_scrub_lock);
 	spa->spa_load_verify_bytes -= BP_GET_PSIZE(bp);
 	cv_broadcast(&spa->spa_scrub_io_cv);
@@ -5353,8 +5256,6 @@ zdb_blkptr_done(zio_t *zio)
 		    blkbuf);
 	}
 	mutex_exit(&spa->spa_scrub_lock);
-
-	abd_free(zio->io_abd);
 }
 
 static int
@@ -5964,7 +5865,6 @@ zdb_leak_init_prepare_indirect_vdevs(spa_t *spa, zdb_cb_t *zcb)
 		 * metaslabs.  We want to set them up for
 		 * zio_claim().
 		 */
-		vdev_metaslab_group_create(vd);
 		VERIFY0(vdev_metaslab_init(vd, 0));
 
 		vdev_indirect_mapping_t *vim __maybe_unused =
@@ -6004,7 +5904,6 @@ zdb_leak_init(spa_t *spa, zdb_cb_t *zcb)
 	 */
 	spa->spa_normal_class->mc_ops = &zdb_metaslab_ops;
 	spa->spa_log_class->mc_ops = &zdb_metaslab_ops;
-	spa->spa_embedded_log_class->mc_ops = &zdb_metaslab_ops;
 
 	zcb->zcb_vd_obsolete_counts =
 	    umem_zalloc(rvd->vdev_children * sizeof (uint32_t *),
@@ -6138,6 +6037,7 @@ zdb_leak_fini(spa_t *spa, zdb_cb_t *zcb)
 	vdev_t *rvd = spa->spa_root_vdev;
 	for (unsigned c = 0; c < rvd->vdev_children; c++) {
 		vdev_t *vd = rvd->vdev_child[c];
+		metaslab_group_t *mg __maybe_unused = vd->vdev_mg;
 
 		if (zcb->zcb_vd_obsolete_counts[c] != NULL) {
 			leaks |= zdb_check_for_obsolete_leaks(vd, zcb);
@@ -6145,9 +6045,7 @@ zdb_leak_fini(spa_t *spa, zdb_cb_t *zcb)
 
 		for (uint64_t m = 0; m < vd->vdev_ms_count; m++) {
 			metaslab_t *msp = vd->vdev_ms[m];
-			ASSERT3P(msp->ms_group, ==, (msp->ms_group->mg_class ==
+			ASSERT3P(mg, ==, msp->ms_group);
-			    spa_embedded_log_class(spa)) ?
-			    vd->vdev_log_mg : vd->vdev_mg);
 
 			/*
 			 * ms_allocatable has been overloaded
@@ -6354,8 +6252,6 @@ dump_block_stats(spa_t *spa)
 	zcb.zcb_totalasize = metaslab_class_get_alloc(spa_normal_class(spa));
 	zcb.zcb_totalasize += metaslab_class_get_alloc(spa_special_class(spa));
 	zcb.zcb_totalasize += metaslab_class_get_alloc(spa_dedup_class(spa));
-	zcb.zcb_totalasize +=
-	    metaslab_class_get_alloc(spa_embedded_log_class(spa));
 	zcb.zcb_start = zcb.zcb_lastprint = gethrtime();
 	err = traverse_pool(spa, 0, flags, zdb_blkptr_cb, &zcb);
 
@@ -6403,7 +6299,6 @@ dump_block_stats(spa_t *spa)
 
 	total_alloc = norm_alloc +
 	    metaslab_class_get_alloc(spa_log_class(spa)) +
-	    metaslab_class_get_alloc(spa_embedded_log_class(spa)) +
 	    metaslab_class_get_alloc(spa_special_class(spa)) +
 	    metaslab_class_get_alloc(spa_dedup_class(spa)) +
 	    get_unflushed_alloc_space(spa);
@@ -6449,7 +6344,7 @@ dump_block_stats(spa_t *spa)
 	(void) printf("\t%-16s %14llu     used: %5.2f%%\n", "Normal class:",
 	    (u_longlong_t)norm_alloc, 100.0 * norm_alloc / norm_space);
 
-	if (spa_special_class(spa)->mc_allocator[0].mca_rotor != NULL) {
+	if (spa_special_class(spa)->mc_rotor != NULL) {
 		uint64_t alloc = metaslab_class_get_alloc(
 		    spa_special_class(spa));
 		uint64_t space = metaslab_class_get_space(
@@ -6460,7 +6355,7 @@ dump_block_stats(spa_t *spa)
 		    100.0 * alloc / space);
 	}
 
-	if (spa_dedup_class(spa)->mc_allocator[0].mca_rotor != NULL) {
+	if (spa_dedup_class(spa)->mc_rotor != NULL) {
 		uint64_t alloc = metaslab_class_get_alloc(
 		    spa_dedup_class(spa));
 		uint64_t space = metaslab_class_get_space(
@@ -6471,17 +6366,6 @@ dump_block_stats(spa_t *spa)
 		    100.0 * alloc / space);
 	}
 
-	if (spa_embedded_log_class(spa)->mc_allocator[0].mca_rotor != NULL) {
-		uint64_t alloc = metaslab_class_get_alloc(
-		    spa_embedded_log_class(spa));
-		uint64_t space = metaslab_class_get_space(
-		    spa_embedded_log_class(spa));
-
-		(void) printf("\t%-16s %14llu     used: %5.2f%%\n",
-		    "Embedded log class", (u_longlong_t)alloc,
-		    100.0 * alloc / space);
-	}
-
 	for (i = 0; i < NUM_BP_EMBEDDED_TYPES; i++) {
 		if (zcb.zcb_embedded_blocks[i] == 0)
 			continue;
@@ -8288,7 +8172,6 @@ main(int argc, char **argv)
 	nvlist_t *policy = NULL;
 	uint64_t max_txg = UINT64_MAX;
 	int64_t objset_id = -1;
-	uint64_t object;
 	int flags = ZFS_IMPORT_MISSING_LOG;
 	int rewind = ZPOOL_NEVER_REWIND;
 	char *spa_config_path_env, *objset_str;
@@ -8317,7 +8200,7 @@ main(int argc, char **argv)
 	zfs_btree_verify_intensity = 3;
 
 	while ((c = getopt(argc, argv,
-	    "AbcCdDeEFGhiI:klLmMo:Op:PqrRsSt:uU:vVx:XYyZ")) != -1) {
+	    "AbcCdDeEFGhiI:klLmMo:Op:PqRsSt:uU:vVx:XYyZ")) != -1) {
 		switch (c) {
 		case 'b':
 		case 'c':
@@ -8332,7 +8215,6 @@ main(int argc, char **argv)
 		case 'm':
 		case 'M':
 		case 'O':
-		case 'r':
 		case 'R':
 		case 's':
 		case 'S':
@@ -8422,7 +8304,7 @@ main(int argc, char **argv)
 		(void) fprintf(stderr, "-p option requires use of -e\n");
 		usage();
 	}
-	if (dump_opt['d'] || dump_opt['r']) {
+	if (dump_opt['d']) {
 		/* <pool>[/<dataset | objset id> is accepted */
 		if (argv[2] && (objset_str = strchr(argv[2], '/')) != NULL &&
 		    objset_str++ != NULL) {
@@ -8475,18 +8357,13 @@ main(int argc, char **argv)
 	 */
 	spa_load_verify_dryrun = B_TRUE;
 
-	/*
-	 * ZDB should have ability to read spacemaps.
-	 */
-	spa_mode_readable_spacemaps = B_TRUE;
-
 	kernel_init(SPA_MODE_READ);
 
 	if (dump_all)
 		verbose = MAX(verbose, 1);
 
 	for (c = 0; c < 256; c++) {
-		if (dump_all && strchr("AeEFklLOPrRSXy", c) == NULL)
+		if (dump_all && strchr("AeEFklLOPRSXy", c) == NULL)
 			dump_opt[c] = 1;
 		if (dump_opt[c])
 			dump_opt[c] += verbose;
@@ -8522,13 +8399,7 @@ main(int argc, char **argv)
 		if (argc != 2)
 			usage();
 		dump_opt['v'] = verbose + 3;
-		return (dump_path(argv[0], argv[1], NULL));
+		return (dump_path(argv[0], argv[1]));
-	}
-	if (dump_opt['r']) {
-		if (argc != 3)
-			usage();
-		dump_opt['v'] = verbose;
-		error = dump_path(argv[0], argv[1], &object);
 	}
 
 	if (dump_opt['X'] || dump_opt['F'])
@@ -8706,9 +8577,7 @@ main(int argc, char **argv)
 
 	argv++;
 	argc--;
-	if (dump_opt['r']) {
+	if (!dump_opt['R']) {
-		error = zdb_copy_object(os, object, argv[1]);
-	} else if (!dump_opt['R']) {
 		flagbits['d'] = ZOR_FLAG_DIRECTORY;
 		flagbits['f'] = ZOR_FLAG_PLAIN_FILE;
 		flagbits['m'] = ZOR_FLAG_SPACE_MAP;

cmd/zed/Makefile.am

@@ -1,10 +1,8 @@
 include $(top_srcdir)/config/Rules.am
-include $(top_srcdir)/config/Shellcheck.am
 
 AM_CFLAGS += $(LIBUDEV_CFLAGS) $(LIBUUID_CFLAGS)
 
 SUBDIRS = zed.d
-SHELLCHECKDIRS = $(SUBDIRS)
 
 sbin_PROGRAMS = zed
 
@@ -45,7 +43,7 @@ zed_LDADD = \
 	$(abs_top_builddir)/lib/libnvpair/libnvpair.la \
 	$(abs_top_builddir)/lib/libuutil/libuutil.la
 
-zed_LDADD += -lrt $(LIBATOMIC_LIBS) $(LIBUDEV_LIBS) $(LIBUUID_LIBS)
+zed_LDADD += -lrt $(LIBUDEV_LIBS) $(LIBUUID_LIBS)
 zed_LDFLAGS = -pthread
 
 EXTRA_DIST = agents/README.md

cmd/zed/agents/zfs_agents.c

@@ -13,7 +13,6 @@
 /*
  * Copyright (c) 2016, Intel Corporation.
  * Copyright (c) 2018, loli10K <ezomori.nozomu@gmail.com>
- * Copyright (c) 2021 Hewlett Packard Enterprise Development LP
  */
 
 #include <libnvpair.h>
@@ -212,18 +211,12 @@ zfs_agent_post_event(const char *class, const char *subclass, nvlist_t *nvl)
 		 * For multipath, spare and l2arc devices ZFS_EV_VDEV_GUID or
 		 * ZFS_EV_POOL_GUID may be missing so find them.
 		 */
-		if (pool_guid == 0 || vdev_guid == 0) {
+		(void) nvlist_lookup_string(nvl, DEV_IDENTIFIER,
-			if ((nvlist_lookup_string(nvl, DEV_IDENTIFIER,
+		    &search.gs_devid);
-			    &search.gs_devid) == 0) &&
+		(void) zpool_iter(g_zfs_hdl, zfs_agent_iter_pool, &search);
-			    (zpool_iter(g_zfs_hdl, zfs_agent_iter_pool, &search)
+		pool_guid = search.gs_pool_guid;
-			    == 1)) {
+		vdev_guid = search.gs_vdev_guid;
-				if (pool_guid == 0)
+		devtype = search.gs_vdev_type;
-					pool_guid = search.gs_pool_guid;
-				if (vdev_guid == 0)
-					vdev_guid = search.gs_vdev_guid;
-				devtype = search.gs_vdev_type;
-			}
-		}
 
 		/*
 		 * We want to avoid reporting "remove" events coming from
@@ -392,7 +385,6 @@ zfs_agent_init(libzfs_handle_t *zfs_hdl)
 		list_destroy(&agent_events);
 		zed_log_die("Failed to initialize agents");
 	}
-	pthread_setname_np(g_agents_tid, "agents");
 }
 
 void

cmd/zed/agents/zfs_diagnosis.c

@@ -35,7 +35,6 @@
 #include <sys/fs/zfs.h>
 #include <sys/fm/protocol.h>
 #include <sys/fm/fs/zfs.h>
-#include <sys/zio.h>
 
 #include "zfs_agents.h"
 #include "fmd_api.h"
@@ -774,8 +773,6 @@ zfs_fm_recv(fmd_hdl_t *hdl, fmd_event_t *ep, nvlist_t *nvl, const char *class)
 	    ZFS_MAKE_EREPORT(FM_EREPORT_ZFS_PROBE_FAILURE))) {
 		char *failmode = NULL;
 		boolean_t checkremove = B_FALSE;
-		uint32_t pri = 0;
-		int32_t flags = 0;
 
 		/*
 		 * If this is a checksum or I/O error, then toss it into the
@@ -798,23 +795,6 @@ zfs_fm_recv(fmd_hdl_t *hdl, fmd_event_t *ep, nvlist_t *nvl, const char *class)
 				checkremove = B_TRUE;
 		} else if (fmd_nvl_class_match(hdl, nvl,
 		    ZFS_MAKE_EREPORT(FM_EREPORT_ZFS_CHECKSUM))) {
-			/*
-			 * We ignore ereports for checksum errors generated by
-			 * scrub/resilver I/O to avoid potentially further
-			 * degrading the pool while it's being repaired.
-			 */
-			if (((nvlist_lookup_uint32(nvl,
-			    FM_EREPORT_PAYLOAD_ZFS_ZIO_PRIORITY, &pri) == 0) &&
-			    (pri == ZIO_PRIORITY_SCRUB ||
-			    pri == ZIO_PRIORITY_REBUILD)) ||
-			    ((nvlist_lookup_int32(nvl,
-			    FM_EREPORT_PAYLOAD_ZFS_ZIO_FLAGS, &flags) == 0) &&
-			    (flags & (ZIO_FLAG_SCRUB | ZIO_FLAG_RESILVER)))) {
-				fmd_hdl_debug(hdl, "ignoring '%s' for "
-				    "scrub/resilver I/O", class);
-				return;
-			}
-
 			if (zcp->zc_data.zc_serd_checksum[0] == '\0') {
 				zfs_serd_name(zcp->zc_data.zc_serd_checksum,
 				    pool_guid, vdev_guid, "checksum");

cmd/zed/agents/zfs_mod.c

@@ -183,14 +183,14 @@ zfs_process_add(zpool_handle_t *zhp, nvlist_t *vdev, boolean_t labeled)
 	nvlist_t *nvroot, *newvd;
 	pendingdev_t *device;
 	uint64_t wholedisk = 0ULL;
-	uint64_t offline = 0ULL, faulted = 0ULL;
+	uint64_t offline = 0ULL;
 	uint64_t guid = 0ULL;
 	char *physpath = NULL, *new_devid = NULL, *enc_sysfs_path = NULL;
 	char rawpath[PATH_MAX], fullpath[PATH_MAX];
 	char devpath[PATH_MAX];
 	int ret;
+	boolean_t is_dm = B_FALSE;
 	boolean_t is_sd = B_FALSE;
-	boolean_t is_mpath_wholedisk = B_FALSE;
 	uint_t c;
 	vdev_stat_t *vs;
 
@@ -211,73 +211,15 @@ zfs_process_add(zpool_handle_t *zhp, nvlist_t *vdev, boolean_t labeled)
 	    &enc_sysfs_path);
 	(void) nvlist_lookup_uint64(vdev, ZPOOL_CONFIG_WHOLE_DISK, &wholedisk);
 	(void) nvlist_lookup_uint64(vdev, ZPOOL_CONFIG_OFFLINE, &offline);
-	(void) nvlist_lookup_uint64(vdev, ZPOOL_CONFIG_FAULTED, &faulted);
-
 	(void) nvlist_lookup_uint64(vdev, ZPOOL_CONFIG_GUID, &guid);
 
-	/*
+	if (offline)
-	 * Special case:
+		return;  /* don't intervene if it was taken offline */
-	 *
-	 * We've seen times where a disk won't have a ZPOOL_CONFIG_PHYS_PATH
-	 * entry in their config. For example, on this force-faulted disk:
-	 *
-	 *	children[0]:
-	 *	   type: 'disk'
-	 *	   id: 0
-	 *	   guid: 14309659774640089719
-	 *        path: '/dev/disk/by-vdev/L28'
-	 *        whole_disk: 0
-	 *        DTL: 654
-	 *        create_txg: 4
-	 *        com.delphix:vdev_zap_leaf: 1161
-	 *        faulted: 1
-	 *        aux_state: 'external'
-	 *	children[1]:
-	 *        type: 'disk'
-	 *        id: 1
-	 *        guid: 16002508084177980912
-	 *        path: '/dev/disk/by-vdev/L29'
-	 *        devid: 'dm-uuid-mpath-35000c500a61d68a3'
-	 *        phys_path: 'L29'
-	 *        vdev_enc_sysfs_path: '/sys/class/enclosure/0:0:1:0/SLOT 30 32'
-	 *        whole_disk: 0
-	 *        DTL: 1028
-	 *        create_txg: 4
-	 *        com.delphix:vdev_zap_leaf: 131
-	 *
-	 * If the disk's path is a /dev/disk/by-vdev/ path, then we can infer
-	 * the ZPOOL_CONFIG_PHYS_PATH from the by-vdev disk name.
-	 */
-	if (physpath == NULL && path != NULL) {
-		/* If path begins with "/dev/disk/by-vdev/" ... */
-		if (strncmp(path, DEV_BYVDEV_PATH,
-		    strlen(DEV_BYVDEV_PATH)) == 0) {
-			/* Set physpath to the char after "/dev/disk/by-vdev" */
-			physpath = &path[strlen(DEV_BYVDEV_PATH)];
-		}
-	}
 
-	/*
+	is_dm = zfs_dev_is_dm(path);
-	 * We don't want to autoreplace offlined disks.  However, we do want to
-	 * replace force-faulted disks (`zpool offline -f`).  Force-faulted
-	 * disks have both offline=1 and faulted=1 in the nvlist.
-	 */
-	if (offline && !faulted) {
-		zed_log_msg(LOG_INFO, "%s: %s is offline, skip autoreplace",
-		    __func__, path);
-		return;
-	}
-
-	is_mpath_wholedisk = is_mpath_whole_disk(path);
 	zed_log_msg(LOG_INFO, "zfs_process_add: pool '%s' vdev '%s', phys '%s'"
-	    " %s blank disk, %s mpath blank disk, %s labeled, enc sysfs '%s', "
+	    " wholedisk %d, %s dm (guid %llu)", zpool_get_name(zhp), path,
-	    "(guid %llu)",
+	    physpath ? physpath : "NULL", wholedisk, is_dm ? "is" : "not",
-	    zpool_get_name(zhp), path,
-	    physpath ? physpath : "NULL",
-	    wholedisk ? "is" : "not",
-	    is_mpath_wholedisk? "is" : "not",
-	    labeled ? "is" : "not",
-	    enc_sysfs_path,
 	    (long long unsigned int)guid);
 
 	/*
@@ -311,9 +253,8 @@ zfs_process_add(zpool_handle_t *zhp, nvlist_t *vdev, boolean_t labeled)
 	    ZFS_ONLINE_CHECKREMOVE | ZFS_ONLINE_UNSPARE, &newstate) == 0 &&
 	    (newstate == VDEV_STATE_HEALTHY ||
 	    newstate == VDEV_STATE_DEGRADED)) {
-		zed_log_msg(LOG_INFO,
+		zed_log_msg(LOG_INFO, "  zpool_vdev_online: vdev %s is %s",
-		    "  zpool_vdev_online: vdev '%s' ('%s') is "
+		    fullpath, (newstate == VDEV_STATE_HEALTHY) ?
-		    "%s", fullpath, physpath, (newstate == VDEV_STATE_HEALTHY) ?
 		    "HEALTHY" : "DEGRADED");
 		return;
 	}
@@ -330,12 +271,11 @@ zfs_process_add(zpool_handle_t *zhp, nvlist_t *vdev, boolean_t labeled)
 	 * vdev online to trigger a FMA fault by posting an ereport.
 	 */
 	if (!zpool_get_prop_int(zhp, ZPOOL_PROP_AUTOREPLACE, NULL) ||
-	    !(wholedisk || is_mpath_wholedisk) || (physpath == NULL)) {
+	    !(wholedisk || is_dm) || (physpath == NULL)) {
 		(void) zpool_vdev_online(zhp, fullpath, ZFS_ONLINE_FORCEFAULT,
 		    &newstate);
 		zed_log_msg(LOG_INFO, "Pool's autoreplace is not enabled or "
-		    "not a blank disk for '%s' ('%s')", fullpath,
+		    "not a whole disk for '%s'", fullpath);
-		    physpath);
 		return;
 	}
 
@@ -347,7 +287,7 @@ zfs_process_add(zpool_handle_t *zhp, nvlist_t *vdev, boolean_t labeled)
 	(void) snprintf(rawpath, sizeof (rawpath), "%s%s",
 	    is_sd ? DEV_BYVDEV_PATH : DEV_BYPATH_PATH, physpath);
 
-	if (realpath(rawpath, devpath) == NULL && !is_mpath_wholedisk) {
+	if (realpath(rawpath, devpath) == NULL && !is_dm) {
 		zed_log_msg(LOG_INFO, "  realpath: %s failed (%s)",
 		    rawpath, strerror(errno));
 
@@ -363,14 +303,12 @@ zfs_process_add(zpool_handle_t *zhp, nvlist_t *vdev, boolean_t labeled)
 	if ((vs->vs_state != VDEV_STATE_DEGRADED) &&
 	    (vs->vs_state != VDEV_STATE_FAULTED) &&
 	    (vs->vs_state != VDEV_STATE_CANT_OPEN)) {
-		zed_log_msg(LOG_INFO, "  not autoreplacing since disk isn't in "
-		    "a bad state (currently %d)", vs->vs_state);
 		return;
 	}
 
 	nvlist_lookup_string(vdev, "new_devid", &new_devid);
 
-	if (is_mpath_wholedisk) {
+	if (is_dm) {
 		/* Don't label device mapper or multipath disks. */
 	} else if (!labeled) {
 		/*
@@ -497,15 +435,7 @@ zfs_process_add(zpool_handle_t *zhp, nvlist_t *vdev, boolean_t labeled)
 		return;
 	}
 
-	/*
+	ret = zpool_vdev_attach(zhp, fullpath, path, nvroot, B_TRUE, B_FALSE);
-	 * Prefer sequential resilvering when supported (mirrors and dRAID),
-	 * otherwise fallback to a traditional healing resilver.
-	 */
-	ret = zpool_vdev_attach(zhp, fullpath, path, nvroot, B_TRUE, B_TRUE);
-	if (ret != 0) {
-		ret = zpool_vdev_attach(zhp, fullpath, path, nvroot,
-		    B_TRUE, B_FALSE);
-	}
 
 	zed_log_msg(LOG_INFO, "  zpool_vdev_replace: %s with %s (%s)",
 	    fullpath, path, (ret == 0) ? "no errors" :
@@ -584,11 +514,8 @@ zfs_iter_vdev(zpool_handle_t *zhp, nvlist_t *nvl, void *data)
 		 * the dp->dd_compare value.
 		 */
 		if (nvlist_lookup_string(nvl, dp->dd_prop, &path) != 0 ||
-		    strcmp(dp->dd_compare, path) != 0) {
+		    strcmp(dp->dd_compare, path) != 0)
-			zed_log_msg(LOG_INFO, "  %s: no match (%s != vdev %s)",
-			    __func__, dp->dd_compare, path);
 			return;
-		}
 
 		zed_log_msg(LOG_INFO, "  zfs_iter_vdev: matched %s on %s",
 		    dp->dd_prop, path);
@@ -636,8 +563,6 @@ zfs_iter_pool(zpool_handle_t *zhp, void *data)
 			    ZPOOL_CONFIG_VDEV_TREE, &nvl);
 			zfs_iter_vdev(zhp, nvl, data);
 		}
-	} else {
-		zed_log_msg(LOG_INFO, "%s: no config\n", __func__);
 	}
 
 	/*
@@ -686,72 +611,6 @@ devphys_iter(const char *physical, const char *devid, zfs_process_func_t func,
 	return (data.dd_found);
 }
 
-/*
- * Given a device identifier, find any vdevs with a matching by-vdev
- * path.  Normally we shouldn't need this as the comparison would be
- * made earlier in the devphys_iter().  For example, if we were replacing
- * /dev/disk/by-vdev/L28, normally devphys_iter() would match the
- * ZPOOL_CONFIG_PHYS_PATH of "L28" from the old disk config to "L28"
- * of the new disk config.  However, we've seen cases where
- * ZPOOL_CONFIG_PHYS_PATH was not in the config for the old disk.  Here's
- * an example of a real 2-disk mirror pool where one disk was force
- * faulted:
- *
- *       com.delphix:vdev_zap_top: 129
- *           children[0]:
- *               type: 'disk'
- *               id: 0
- *               guid: 14309659774640089719
- *               path: '/dev/disk/by-vdev/L28'
- *               whole_disk: 0
- *               DTL: 654
- *               create_txg: 4
- *               com.delphix:vdev_zap_leaf: 1161
- *               faulted: 1
- *               aux_state: 'external'
- *           children[1]:
- *               type: 'disk'
- *               id: 1
- *               guid: 16002508084177980912
- *               path: '/dev/disk/by-vdev/L29'
- *               devid: 'dm-uuid-mpath-35000c500a61d68a3'
- *               phys_path: 'L29'
- *               vdev_enc_sysfs_path: '/sys/class/enclosure/0:0:1:0/SLOT 30 32'
- *               whole_disk: 0
- *               DTL: 1028
- *               create_txg: 4
- *               com.delphix:vdev_zap_leaf: 131
- *
- * So in the case above, the only thing we could compare is the path.
- *
- * We can do this because we assume by-vdev paths are authoritative as physical
- * paths.  We could not assume this for normal paths like /dev/sda since the
- * physical location /dev/sda points to could change over time.
- */
-static boolean_t
-by_vdev_path_iter(const char *by_vdev_path, const char *devid,
-    zfs_process_func_t func, boolean_t is_slice)
-{
-	dev_data_t data = { 0 };
-
-	data.dd_compare = by_vdev_path;
-	data.dd_func = func;
-	data.dd_prop = ZPOOL_CONFIG_PATH;
-	data.dd_found = B_FALSE;
-	data.dd_islabeled = is_slice;
-	data.dd_new_devid = devid;
-
-	if (strncmp(by_vdev_path, DEV_BYVDEV_PATH,
-	    strlen(DEV_BYVDEV_PATH)) != 0) {
-		/* by_vdev_path doesn't start with "/dev/disk/by-vdev/" */
-		return (B_FALSE);
-	}
-
-	(void) zpool_iter(g_zfshdl, zfs_iter_pool, &data);
-
-	return (data.dd_found);
-}
-
 /*
  * Given a device identifier, find any vdevs with a matching devid.
  * On Linux we can match devid directly which is always a whole disk.
@@ -773,27 +632,6 @@ devid_iter(const char *devid, zfs_process_func_t func, boolean_t is_slice)
 	return (data.dd_found);
 }
 
-/*
- * Given a device guid, find any vdevs with a matching guid.
- */
-static boolean_t
-guid_iter(uint64_t pool_guid, uint64_t vdev_guid, const char *devid,
-    zfs_process_func_t func, boolean_t is_slice)
-{
-	dev_data_t data = { 0 };
-
-	data.dd_func = func;
-	data.dd_found = B_FALSE;
-	data.dd_pool_guid = pool_guid;
-	data.dd_vdev_guid = vdev_guid;
-	data.dd_islabeled = is_slice;
-	data.dd_new_devid = devid;
-
-	(void) zpool_iter(g_zfshdl, zfs_iter_pool, &data);
-
-	return (data.dd_found);
-}
-
 /*
  * Handle a EC_DEV_ADD.ESC_DISK event.
  *
@@ -816,21 +654,16 @@ guid_iter(uint64_t pool_guid, uint64_t vdev_guid, const char *devid,
 static int
 zfs_deliver_add(nvlist_t *nvl, boolean_t is_lofi)
 {
-	char *devpath = NULL, *devid = NULL;
+	char *devpath = NULL, *devid;
-	uint64_t pool_guid = 0, vdev_guid = 0;
 	boolean_t is_slice;
 
 	/*
-	 * Expecting a devid string and an optional physical location and guid
+	 * Expecting a devid string and an optional physical location
 	 */
-	if (nvlist_lookup_string(nvl, DEV_IDENTIFIER, &devid) != 0) {
+	if (nvlist_lookup_string(nvl, DEV_IDENTIFIER, &devid) != 0)
-		zed_log_msg(LOG_INFO, "%s: no dev identifier\n", __func__);
 		return (-1);
-	}
 
 	(void) nvlist_lookup_string(nvl, DEV_PHYS_PATH, &devpath);
-	(void) nvlist_lookup_uint64(nvl, ZFS_EV_POOL_GUID, &pool_guid);
-	(void) nvlist_lookup_uint64(nvl, ZFS_EV_VDEV_GUID, &vdev_guid);
 
 	is_slice = (nvlist_lookup_boolean(nvl, DEV_IS_PART) == 0);
 
@@ -841,28 +674,12 @@ zfs_deliver_add(nvlist_t *nvl, boolean_t is_lofi)
 	 * Iterate over all vdevs looking for a match in the following order:
 	 * 1. ZPOOL_CONFIG_DEVID (identifies the unique disk)
 	 * 2. ZPOOL_CONFIG_PHYS_PATH (identifies disk physical location).
-	 * 3. ZPOOL_CONFIG_GUID (identifies unique vdev).
+	 *
-	 * 4. ZPOOL_CONFIG_PATH for /dev/disk/by-vdev devices only (since
+	 * For disks, we only want to pay attention to vdevs marked as whole
-	 *    by-vdev paths represent physical paths).
+	 * disks or are a multipath device.
 	 */
-	if (devid_iter(devid, zfs_process_add, is_slice))
+	if (!devid_iter(devid, zfs_process_add, is_slice) && devpath != NULL)
-		return (0);
+		(void) devphys_iter(devpath, devid, zfs_process_add, is_slice);
-	if (devpath != NULL && devphys_iter(devpath, devid, zfs_process_add,
-	    is_slice))
-		return (0);
-	if (vdev_guid != 0)
-		(void) guid_iter(pool_guid, vdev_guid, devid, zfs_process_add,
-		    is_slice);
-
-	if (devpath != NULL) {
-		/* Can we match a /dev/disk/by-vdev/ path? */
-		char by_vdev_path[MAXPATHLEN];
-		snprintf(by_vdev_path, sizeof (by_vdev_path),
-		    "/dev/disk/by-vdev/%s", devpath);
-		if (by_vdev_path_iter(by_vdev_path, devid, zfs_process_add,
-		    is_slice))
-			return (0);
-	}
 
 	return (0);
 }
@@ -1093,7 +910,6 @@ zfs_slm_init()
 		return (-1);
 	}
 
-	pthread_setname_np(g_zfs_tid, "enum-pools");
 	list_create(&g_device_list, sizeof (struct pendingdev),
 	    offsetof(struct pendingdev, pd_node));
 

cmd/zed/agents/zfs_retire.c

@@ -40,7 +40,6 @@
 #include <sys/fm/fs/zfs.h>
 #include <libzfs.h>
 #include <string.h>
-#include <libgen.h>
 
 #include "zfs_agents.h"
 #include "fmd_api.h"
@@ -220,18 +219,12 @@ replace_with_spare(fmd_hdl_t *hdl, zpool_handle_t *zhp, nvlist_t *vdev)
 	 * replace it.
 	 */
 	for (s = 0; s < nspares; s++) {
-		boolean_t rebuild = B_FALSE;
+		char *spare_name;
-		char *spare_name, *type;
 
 		if (nvlist_lookup_string(spares[s], ZPOOL_CONFIG_PATH,
 		    &spare_name) != 0)
 			continue;
 
-		/* prefer sequential resilvering for distributed spares */
-		if ((nvlist_lookup_string(spares[s], ZPOOL_CONFIG_TYPE,
-		    &type) == 0) && strcmp(type, VDEV_TYPE_DRAID_SPARE) == 0)
-			rebuild = B_TRUE;
-
 		/* if set, add the "ashift" pool property to the spare nvlist */
 		if (source != ZPROP_SRC_DEFAULT)
 			(void) nvlist_add_uint64(spares[s],
@@ -244,7 +237,7 @@ replace_with_spare(fmd_hdl_t *hdl, zpool_handle_t *zhp, nvlist_t *vdev)
 		    dev_name, basename(spare_name));
 
 		if (zpool_vdev_attach(zhp, dev_name, spare_name,
-		    replacement, B_TRUE, rebuild) == 0) {
+		    replacement, B_TRUE, B_FALSE) == 0) {
 			free(dev_name);
 			nvlist_free(replacement);
 			return (B_TRUE);
@@ -335,7 +328,7 @@ zfs_retire_recv(fmd_hdl_t *hdl, fmd_event_t *ep, nvlist_t *nvl,
 	 */
 	if (strcmp(class, "resource.fs.zfs.removed") == 0 ||
 	    (strcmp(class, "resource.fs.zfs.statechange") == 0 &&
-	    (state == VDEV_STATE_REMOVED || state == VDEV_STATE_FAULTED))) {
+	    state == VDEV_STATE_REMOVED)) {
 		char *devtype;
 		char *devname;
 
@@ -506,7 +499,6 @@ zfs_retire_recv(fmd_hdl_t *hdl, fmd_event_t *ep, nvlist_t *nvl,
 		 * Attempt to substitute a hot spare.
 		 */
 		(void) replace_with_spare(hdl, zhp, vdev);
-
 		zpool_close(zhp);
 	}
 

cmd/zed/zed.c

@@ -3,7 +3,7 @@
  *
  * Developed at Lawrence Livermore National Laboratory (LLNL-CODE-403049).
  * Copyright (C) 2013-2014 Lawrence Livermore National Security, LLC.
- * Refer to the OpenZFS git commit log for authoritative copyright attribution.
+ * Refer to the ZoL git commit log for authoritative copyright attribution.
  *
  * The contents of this file are subject to the terms of the
  * Common Development and Distribution License Version 1.0 (CDDL-1.0).
@@ -60,8 +60,8 @@ _setup_sig_handlers(void)
 		zed_log_die("Failed to initialize sigset");
 
 	sa.sa_flags = SA_RESTART;
-
 	sa.sa_handler = SIG_IGN;
+
 	if (sigaction(SIGPIPE, &sa, NULL) < 0)
 		zed_log_die("Failed to ignore SIGPIPE");
 
@@ -75,10 +75,6 @@ _setup_sig_handlers(void)
 	sa.sa_handler = _hup_handler;
 	if (sigaction(SIGHUP, &sa, NULL) < 0)
 		zed_log_die("Failed to register SIGHUP handler");
-
-	(void) sigaddset(&sa.sa_mask, SIGCHLD);
-	if (pthread_sigmask(SIG_BLOCK, &sa.sa_mask, NULL) < 0)
-		zed_log_die("Failed to block SIGCHLD");
 }
 
 /*
@@ -216,20 +212,22 @@ _finish_daemonize(void)
 int
 main(int argc, char *argv[])
 {
-	struct zed_conf zcp;
+	struct zed_conf *zcp;
 	uint64_t saved_eid;
 	int64_t saved_etime[2];
 
 	zed_log_init(argv[0]);
 	zed_log_stderr_open(LOG_NOTICE);
-	zed_conf_init(&zcp);
+	zcp = zed_conf_create();
-	zed_conf_parse_opts(&zcp, argc, argv);
+	zed_conf_parse_opts(zcp, argc, argv);
-	if (zcp.do_verbose)
+	if (zcp->do_verbose)
 		zed_log_stderr_open(LOG_INFO);
 
 	if (geteuid() != 0)
 		zed_log_die("Must be run as root");
 
+	zed_conf_parse_file(zcp);
+
 	zed_file_close_from(STDERR_FILENO + 1);
 
 	(void) umask(0);
@@ -237,32 +235,32 @@ main(int argc, char *argv[])
 	if (chdir("/") < 0)
 		zed_log_die("Failed to change to root directory");
 
-	if (zed_conf_scan_dir(&zcp) < 0)
+	if (zed_conf_scan_dir(zcp) < 0)
 		exit(EXIT_FAILURE);
 
-	if (!zcp.do_foreground) {
+	if (!zcp->do_foreground) {
 		_start_daemonize();
 		zed_log_syslog_open(LOG_DAEMON);
 	}
 	_setup_sig_handlers();
 
-	if (zcp.do_memlock)
+	if (zcp->do_memlock)
 		_lock_memory();
 
-	if ((zed_conf_write_pid(&zcp) < 0) && (!zcp.do_force))
+	if ((zed_conf_write_pid(zcp) < 0) && (!zcp->do_force))
 		exit(EXIT_FAILURE);
 
-	if (!zcp.do_foreground)
+	if (!zcp->do_foreground)
 		_finish_daemonize();
 
 	zed_log_msg(LOG_NOTICE,
 	    "ZFS Event Daemon %s-%s (PID %d)",
 	    ZFS_META_VERSION, ZFS_META_RELEASE, (int)getpid());
 
-	if (zed_conf_open_state(&zcp) < 0)
+	if (zed_conf_open_state(zcp) < 0)
 		exit(EXIT_FAILURE);
 
-	if (zed_conf_read_state(&zcp, &saved_eid, saved_etime) < 0)
+	if (zed_conf_read_state(zcp, &saved_eid, saved_etime) < 0)
 		exit(EXIT_FAILURE);
 
 idle:
@@ -271,38 +269,38 @@ idle:
 	 * successful.
 	 */
 	do {
-		if (!zed_event_init(&zcp))
+		if (!zed_event_init(zcp))
 			break;
 		/* Wait for some time and try again. tunable? */
 		sleep(30);
-	} while (!_got_exit && zcp.do_idle);
+	} while (!_got_exit && zcp->do_idle);
 
 	if (_got_exit)
 		goto out;
 
-	zed_event_seek(&zcp, saved_eid, saved_etime);
+	zed_event_seek(zcp, saved_eid, saved_etime);
 
 	while (!_got_exit) {
 		int rv;
 		if (_got_hup) {
 			_got_hup = 0;
-			(void) zed_conf_scan_dir(&zcp);
+			(void) zed_conf_scan_dir(zcp);
 		}
-		rv = zed_event_service(&zcp);
+		rv = zed_event_service(zcp);
 
 		/* ENODEV: When kernel module is unloaded (osx) */
-		if (rv != 0)
+		if (rv == ENODEV)
 			break;
 	}
 
 	zed_log_msg(LOG_NOTICE, "Exiting");
-	zed_event_fini(&zcp);
+	zed_event_fini(zcp);
 
-	if (zcp.do_idle && !_got_exit)
+	if (zcp->do_idle && !_got_exit)
 		goto idle;
 
 out:
-	zed_conf_destroy(&zcp);
+	zed_conf_destroy(zcp);
 	zed_log_fini();
 	exit(EXIT_SUCCESS);
 }

cmd/zed/zed.d/Makefile.am

@@ -1,6 +1,5 @@
 include $(top_srcdir)/config/Rules.am
 include $(top_srcdir)/config/Substfiles.am
-include $(top_srcdir)/config/Shellcheck.am
 
 EXTRA_DIST += README
 
@@ -52,6 +51,3 @@ install-data-hook:
 	    ln -s "$(zedexecdir)/$${f}" "$(DESTDIR)$(zedconfdir)"; \
 	done
 	chmod 0600 "$(DESTDIR)$(zedconfdir)/zed.rc"
-
-# False positive: 1>&"${ZED_FLOCK_FD}" looks suspiciously similar to a >&filename bash extension
-CHECKBASHISMS_IGNORE = -e 'should be >word 2>&1' -e '&"$${ZED_FLOCK_FD}"'

cmd/zed/zed.d/all-debug.sh

@@ -12,11 +12,15 @@
 
 zed_exit_if_ignoring_this_event
 
-zed_lock "${ZED_DEBUG_LOG}"
+lockfile="$(basename -- "${ZED_DEBUG_LOG}").lock"
-{
-	printenv | sort
-	echo
-} 1>&"${ZED_FLOCK_FD}"
-zed_unlock "${ZED_DEBUG_LOG}"
 
+umask 077
+zed_lock "${lockfile}"
+exec >> "${ZED_DEBUG_LOG}"
+
+printenv | sort
+echo
+
+exec >&-
+zed_unlock "${lockfile}"
 exit 0

cmd/zed/zed.d/all-syslog.sh

@@ -21,7 +21,7 @@ if [ "${ZED_SYSLOG_DISPLAY_GUIDS}" = "1" ]; then
     [ -n "${ZEVENT_VDEV_GUID}" ] && msg="${msg} vdev_guid=${ZEVENT_VDEV_GUID}"
 else
     [ -n "${ZEVENT_POOL}" ] && msg="${msg} pool='${ZEVENT_POOL}'"
-    [ -n "${ZEVENT_VDEV_PATH}" ] && msg="${msg} vdev=${ZEVENT_VDEV_PATH##*/}"
+    [ -n "${ZEVENT_VDEV_PATH}" ] && msg="${msg} vdev=$(basename "${ZEVENT_VDEV_PATH}")"
 fi
 
 # log pool state if state is anything other than 'ACTIVE'
@@ -42,7 +42,6 @@ fi
     msg="${msg} delay=$((ZEVENT_ZIO_DELAY / 1000000))ms"
 
 # list the bookmark data together
-# shellcheck disable=SC2153
 [ -n "${ZEVENT_ZIO_OBJSET}" ] && \
     msg="${msg} bookmark=${ZEVENT_ZIO_OBJSET}:${ZEVENT_ZIO_OBJECT}:${ZEVENT_ZIO_LEVEL}:${ZEVENT_ZIO_BLKID}"
 

cmd/zed/zed.d/data-notify.sh

@@ -25,7 +25,7 @@ zed_rate_limit "${rate_limit_tag}" || exit 3
 
 umask 077
 note_subject="ZFS ${ZEVENT_SUBCLASS} error for ${ZEVENT_POOL} on $(hostname)"
-note_pathname="$(mktemp)"
+note_pathname="${TMPDIR:="/tmp"}/$(basename -- "$0").${ZEVENT_EID}.$$"
 {
     echo "ZFS has detected a data error:"
     echo

cmd/zed/zed.d/generic-notify.sh

@@ -23,7 +23,7 @@
 
 # Rate-limit the notification based in part on the filename.
 #
-rate_limit_tag="${ZEVENT_POOL};${ZEVENT_SUBCLASS};${0##*/}"
+rate_limit_tag="${ZEVENT_POOL};${ZEVENT_SUBCLASS};$(basename -- "$0")"
 rate_limit_interval="${ZED_NOTIFY_INTERVAL_SECS}"
 zed_rate_limit "${rate_limit_tag}" "${rate_limit_interval}" || exit 3
 
@@ -31,7 +31,7 @@ umask 077
 pool_str="${ZEVENT_POOL:+" for ${ZEVENT_POOL}"}"
 host_str=" on $(hostname)"
 note_subject="ZFS ${ZEVENT_SUBCLASS} event${pool_str}${host_str}"
-note_pathname="$(mktemp)"
+note_pathname="${TMPDIR:="/tmp"}/$(basename -- "$0").${ZEVENT_EID}.$$"
 {
     echo "ZFS has posted the following event:"
     echo

cmd/zed/zed.d/history_event-zfs-list-cacher.sh.in

@@ -3,8 +3,9 @@
 # Track changes to enumerated pools for use in early-boot
 set -ef
 
-FSLIST="@sysconfdir@/zfs/zfs-list.cache/${ZEVENT_POOL}"
+FSLIST_DIR="@sysconfdir@/zfs/zfs-list.cache"
-FSLIST_TMP="@runstatedir@/zfs-list.cache@${ZEVENT_POOL}"
+FSLIST_TMP="@runstatedir@/zfs-list.cache.new"
+FSLIST="${FSLIST_DIR}/${ZEVENT_POOL}"
 
 # If the pool specific cache file is not writeable, abort
 [ -w "${FSLIST}" ] || exit 0
@@ -13,20 +14,20 @@ FSLIST_TMP="@runstatedir@/zfs-list.cache@${ZEVENT_POOL}"
 . "${ZED_ZEDLET_DIR}/zed-functions.sh"
 
 [ "$ZEVENT_SUBCLASS" != "history_event" ] && exit 0
-zed_check_cmd "${ZFS}" sort diff
+zed_check_cmd "${ZFS}" sort diff grep
 
 # If we are acting on a snapshot, we have nothing to do
-[ "${ZEVENT_HISTORY_DSNAME%@*}" = "${ZEVENT_HISTORY_DSNAME}" ] || exit 0
+printf '%s' "${ZEVENT_HISTORY_DSNAME}" | grep '@' && exit 0
 
-# We lock the output file to avoid simultaneous writes.
+# We obtain a lock on zfs-list to avoid any simultaneous writes.
 # If we run into trouble, log and drop the lock
 abort_alter() {
-  zed_log_msg "Error updating zfs-list.cache for ${ZEVENT_POOL}!"
+  zed_log_msg "Error updating zfs-list.cache!"
-  zed_unlock "${FSLIST}"
+  zed_unlock zfs-list
 }
 
 finished() {
-  zed_unlock "${FSLIST}"
+  zed_unlock zfs-list
   trap - EXIT
   exit 0
 }
@@ -36,7 +37,7 @@ case "${ZEVENT_HISTORY_INTERNAL_NAME}" in
       ;;
 
     export)
-        zed_lock "${FSLIST}"
+        zed_lock zfs-list
         trap abort_alter EXIT
         echo > "${FSLIST}"
         finished
@@ -62,7 +63,7 @@ case "${ZEVENT_HISTORY_INTERNAL_NAME}" in
       ;;
 esac
 
-zed_lock "${FSLIST}"
+zed_lock zfs-list
 trap abort_alter EXIT
 
 PROPS="name,mountpoint,canmount,atime,relatime,devices,exec\
@@ -78,7 +79,7 @@ PROPS="name,mountpoint,canmount,atime,relatime,devices,exec\
 sort "${FSLIST_TMP}" -o "${FSLIST_TMP}"
 
 # Don't modify the file if it hasn't changed
-diff -q "${FSLIST_TMP}" "${FSLIST}" || cat "${FSLIST_TMP}" > "${FSLIST}"
+diff -q "${FSLIST_TMP}" "${FSLIST}" || mv "${FSLIST_TMP}" "${FSLIST}"
 rm -f "${FSLIST_TMP}"
 
 finished

cmd/zed/zed.d/scrub_finish-notify.sh

@@ -41,7 +41,7 @@ fi
 
 umask 077
 note_subject="ZFS ${ZEVENT_SUBCLASS} event for ${ZEVENT_POOL} on $(hostname)"
-note_pathname="$(mktemp)"
+note_pathname="${TMPDIR:="/tmp"}/$(basename -- "$0").${ZEVENT_EID}.$$"
 {
     echo "ZFS has finished a ${action}:"
     echo

cmd/zed/zed.d/statechange-led.sh

@@ -1,21 +1,21 @@
 #!/bin/sh
 #
-# Turn off/on vdevs' enclosure fault LEDs when their pool's state changes.
+# Turn off/on the VDEV's enclosure fault LEDs when the pool's state changes.
 #
-# Turn a vdev's fault LED on if it becomes FAULTED, DEGRADED or UNAVAIL.
+# Turn the VDEV's fault LED on if it becomes FAULTED, DEGRADED or UNAVAIL.
-# Turn its LED off when it's back ONLINE again.
+# Turn the LED off when it's back ONLINE again.
 #
 # This script run in two basic modes:
 #
 # 1. If $ZEVENT_VDEV_ENC_SYSFS_PATH and $ZEVENT_VDEV_STATE_STR are set, then
-# only set the LED for that particular vdev. This is the case for statechange
+# only set the LED for that particular VDEV. This is the case for statechange
 # events and some vdev_* events.
 #
-# 2. If those vars are not set, then check the state of all vdevs in the pool
+# 2. If those vars are not set, then check the state of all VDEVs in the pool
 # and set the LEDs accordingly.  This is the case for pool_import events.
 #
 # Note that this script requires that your enclosure be supported by the
-# Linux SCSI Enclosure services (SES) driver.  The script will do nothing
+# Linux SCSI enclosure services (ses) driver.  The script will do nothing
 # if you have no enclosure, or if your enclosure isn't supported.
 #
 # Exit codes:
@@ -29,8 +29,7 @@
 [ -f "${ZED_ZEDLET_DIR}/zed.rc" ] && . "${ZED_ZEDLET_DIR}/zed.rc"
 . "${ZED_ZEDLET_DIR}/zed-functions.sh"
 
-if [ ! -d /sys/class/enclosure ] && [ ! -d /sys/bus/pci/slots ] ; then
+if [ ! -d /sys/class/enclosure ] ; then
-	# No JBOD enclosure or NVMe slots
 	exit 1
 fi
 
@@ -60,10 +59,6 @@ check_and_set_led()
 	file="$1"
 	val="$2"
 
-	if [ -z "$val" ]; then
-		return 0
-	fi
-
 	if [ ! -e "$file" ] ; then
 		return 3
 	fi
@@ -71,11 +66,11 @@ check_and_set_led()
 	# If another process is accessing the LED when we attempt to update it,
 	# the update will be lost so retry until the LED actually changes or we
 	# timeout.
-	for _ in 1 2 3 4 5; do
+	for _ in $(seq 1 5); do
 		# We want to check the current state first, since writing to the
 		# 'fault' entry always causes a SES command, even if the
 		# current state is already what you want.
-		read -r current < "${file}"
+		current=$(cat "${file}")
 
 		# On some enclosures if you write 1 to fault, and read it back,
 		# it will return 2.  Treat all non-zero values as 1 for
@@ -90,84 +85,27 @@ check_and_set_led()
 		else
 			break
 		fi
-	done
+        done
-}
-
-# Fault LEDs for JBODs and NVMe drives are handled a little differently.
-#
-# On JBODs the fault LED is called 'fault' and on a path like this:
-#
-#   /sys/class/enclosure/0:0:1:0/SLOT 10/fault
-#
-# On NVMe it's called 'attention' and on a path like this:
-#
-#   /sys/bus/pci/slot/0/attention
-#
-# This function returns the full path to the fault LED file for a given
-# enclosure/slot directory.
-#
-path_to_led()
-{
-	dir=$1
-	if [ -f "$dir/fault" ] ; then
-		echo "$dir/fault"
-	elif [ -f "$dir/attention" ] ; then
-		echo "$dir/attention"
-	fi
 }
 
 state_to_val()
 {
 	state="$1"
-	case "$state" in
+	if [ "$state" = "FAULTED" ] || [ "$state" = "DEGRADED" ] || \
-		FAULTED|DEGRADED|UNAVAIL)
+	   [ "$state" = "UNAVAIL" ] ; then
-			echo 1
+		echo 1
-			;;
+	elif [ "$state" = "ONLINE" ] ; then
-		ONLINE)
+		echo 0
-			echo 0
+	fi
-			;;
-	esac
 }
 
+# process_pool ([pool])
 #
-# Given a nvme name like 'nvme0n1', pass back its slot directory
+# Iterate through a pool (or pools) and set the VDEV's enclosure slot LEDs to
-# like "/sys/bus/pci/slots/0"
+# the VDEV's state.
-#
-nvme_dev_to_slot()
-{
-	dev="$1"
-
-	# Get the address "0000:01:00.0"
-	address=$(cat "/sys/class/block/$dev/device/address")
-
-	# For each /sys/bus/pci/slots subdir that is an actual number
-	# (rather than weird directories like "1-3/").
-	# shellcheck disable=SC2010
-	for i in $(ls /sys/bus/pci/slots/ | grep -E "^[0-9]+$") ; do
-		this_address=$(cat "/sys/bus/pci/slots/$i/address")
-
-		# The format of address is a little different between
-		# /sys/class/block/$dev/device/address and
-		# /sys/bus/pci/slots/
-		#
-		# address=           "0000:01:00.0"
-		# this_address =     "0000:01:00"
-		#
-		if echo "$address" | grep -Eq ^"$this_address" ; then
-			echo "/sys/bus/pci/slots/$i"
-			break
-		fi
-	done
-}
-
-
-# process_pool (pool)
-#
-# Iterate through a pool and set the vdevs' enclosure slot LEDs to
-# those vdevs' state.
 #
 # Arguments
-#   pool:	Pool name.
+#   pool:	Optional pool name.  If not specified, iterate though all pools.
 #
 # Return
 #  0 on success, 3 on missing sysfs path
@@ -175,27 +113,19 @@ nvme_dev_to_slot()
 process_pool()
 {
 	pool="$1"
-
-	# The output will be the vdevs only (from "grep '/dev/'"):
-	#
-	#    U45     ONLINE       0     0     0   /dev/sdk          0
-	#    U46     ONLINE       0     0     0   /dev/sdm          0
-	#    U47     ONLINE       0     0     0   /dev/sdn          0
-	#    U50     ONLINE       0     0     0  /dev/sdbn          0
-	#
-	ZPOOL_SCRIPTS_AS_ROOT=1 $ZPOOL status -c upath,fault_led "$pool" | grep '/dev/' | (
 	rc=0
-	while read -r vdev state _ _ _ therest; do
-		# Read out current LED value and path
-		# Get dev name (like 'sda')
-		dev=$(basename "$(echo "$therest" | awk '{print $(NF-1)}')")
-		vdev_enc_sysfs_path=$(realpath "/sys/class/block/$dev/device/enclosure_device"*)
-		if [ ! -d "$vdev_enc_sysfs_path" ] ; then
-			# This is not a JBOD disk, but it could be a PCI NVMe drive
-			vdev_enc_sysfs_path=$(nvme_dev_to_slot "$dev")
-		fi
 
-		current_val=$(echo "$therest" | awk '{print $NF}')
+	# Lookup all the current LED values and paths in parallel
+	#shellcheck disable=SC2016
+	cmd='echo led_token=$(cat "$VDEV_ENC_SYSFS_PATH/fault"),"$VDEV_ENC_SYSFS_PATH",'
+	out=$($ZPOOL status -vc "$cmd" "$pool" | grep 'led_token=')
+
+	#shellcheck disable=SC2034
+	echo "$out" | while read -r vdev state read write chksum therest; do
+		# Read out current LED value and path
+		tmp=$(echo "$therest" | sed 's/^.*led_token=//g')
+		vdev_enc_sysfs_path=$(echo "$tmp" | awk -F ',' '{print $2}')
+		current_val=$(echo "$tmp" | awk -F ',' '{print $1}')
 
 		if [ "$current_val" != "0" ] ; then
 			current_val=1
@@ -206,33 +136,40 @@ process_pool()
 			continue
 		fi
 
-		led_path=$(path_to_led "$vdev_enc_sysfs_path")
+		if [ ! -e "$vdev_enc_sysfs_path/fault" ] ; then
-		if [ ! -e "$led_path" ] ; then
+			#shellcheck disable=SC2030
-			rc=3
+			rc=1
-			zed_log_msg "vdev $vdev '$led_path' doesn't exist"
+			zed_log_msg "vdev $vdev '$file/fault' doesn't exist"
-			continue
+			continue;
 		fi
 
 		val=$(state_to_val "$state")
 
 		if [ "$current_val" = "$val" ] ; then
 			# LED is already set correctly
-			continue
+			continue;
 		fi
 
-		if ! check_and_set_led "$led_path" "$val"; then
+		if ! check_and_set_led "$vdev_enc_sysfs_path/fault" "$val"; then
-			rc=3
+			rc=1
 		fi
+
 	done
-	exit "$rc"; )
+
+	#shellcheck disable=SC2031
+	if [ "$rc" = "0" ] ; then
+		return 0
+	else
+		# We didn't see a sysfs entry that we wanted to set
+		return 3
+	fi
 }
 
 if [ -n "$ZEVENT_VDEV_ENC_SYSFS_PATH" ] && [ -n "$ZEVENT_VDEV_STATE_STR" ] ; then
-	# Got a statechange for an individual vdev
+	# Got a statechange for an individual VDEV
 	val=$(state_to_val "$ZEVENT_VDEV_STATE_STR")
 	vdev=$(basename "$ZEVENT_VDEV_PATH")
-	ledpath=$(path_to_led "$ZEVENT_VDEV_ENC_SYSFS_PATH")
+	check_and_set_led "$ZEVENT_VDEV_ENC_SYSFS_PATH/fault" "$val"
-	check_and_set_led "$ledpath" "$val"
 else
 	# Process the entire pool
 	poolname=$(zed_guid_to_pool "$ZEVENT_POOL_GUID")

cmd/zed/zed.d/statechange-notify.sh

@@ -15,7 +15,7 @@
 # Send notification in response to a fault induced statechange
 #
 # ZEVENT_SUBCLASS: 'statechange'
-# ZEVENT_VDEV_STATE_STR: 'DEGRADED', 'FAULTED', 'REMOVED', or 'UNAVAIL'
+# ZEVENT_VDEV_STATE_STR: 'DEGRADED', 'FAULTED' or 'REMOVED'
 #
 # Exit codes:
 #   0: notification sent
@@ -31,14 +31,13 @@
 
 if [ "${ZEVENT_VDEV_STATE_STR}" != "FAULTED" ] \
         && [ "${ZEVENT_VDEV_STATE_STR}" != "DEGRADED" ] \
-        && [ "${ZEVENT_VDEV_STATE_STR}" != "REMOVED" ] \
+        && [ "${ZEVENT_VDEV_STATE_STR}" != "REMOVED" ]; then
-        && [ "${ZEVENT_VDEV_STATE_STR}" != "UNAVAIL" ]; then
     exit 3
 fi
 
 umask 077
 note_subject="ZFS device fault for pool ${ZEVENT_POOL_GUID} on $(hostname)"
-note_pathname="$(mktemp)"
+note_pathname="${TMPDIR:="/tmp"}/$(basename -- "$0").${ZEVENT_EID}.$$"
 {
     if [ "${ZEVENT_VDEV_STATE_STR}" = "FAULTED" ] ; then
         echo "The number of I/O errors associated with a ZFS device exceeded"

cmd/zed/zed.d/trim_finish-notify.sh

@@ -19,7 +19,7 @@ zed_check_cmd "${ZPOOL}" || exit 9
 
 umask 077
 note_subject="ZFS ${ZEVENT_SUBCLASS} event for ${ZEVENT_POOL} on $(hostname)"
-note_pathname="$(mktemp)"
+note_pathname="${TMPDIR:="/tmp"}/$(basename -- "$0").${ZEVENT_EID}.$$"
 {
     echo "ZFS has finished a trim:"
     echo

cmd/zed/zed.d/zed-functions.sh

@@ -77,7 +77,7 @@ zed_log_msg()
 zed_log_err()
 {
     logger -p "${ZED_SYSLOG_PRIORITY}" -t "${ZED_SYSLOG_TAG}" -- "error:" \
-        "${0##*/}:""${ZEVENT_EID:+" eid=${ZEVENT_EID}:"}" "$@"
+        "$(basename -- "$0"):""${ZEVENT_EID:+" eid=${ZEVENT_EID}:"}" "$@"
 }
 
 
@@ -126,8 +126,10 @@ zed_lock()
 
     # Obtain a lock on the file bound to the given file descriptor.
     #
-    eval "exec ${fd}>> '${lockfile}'"
+    eval "exec ${fd}> '${lockfile}'"
-    if ! err="$(flock --exclusive "${fd}" 2>&1)"; then
+    err="$(flock --exclusive "${fd}" 2>&1)"
+    # shellcheck disable=SC2181
+    if [ $? -ne 0 ]; then
         zed_log_err "failed to lock \"${lockfile}\": ${err}"
     fi
 
@@ -163,7 +165,9 @@ zed_unlock()
     fi
 
     # Release the lock and close the file descriptor.
-    if ! err="$(flock --unlock "${fd}" 2>&1)"; then
+    err="$(flock --unlock "${fd}" 2>&1)"
+    # shellcheck disable=SC2181
+    if [ $? -ne 0 ]; then
         zed_log_err "failed to unlock \"${lockfile}\": ${err}"
     fi
     eval "exec ${fd}>&-"
@@ -202,10 +206,6 @@ zed_notify()
     [ "${rv}" -eq 0 ] && num_success=$((num_success + 1))
     [ "${rv}" -eq 1 ] && num_failure=$((num_failure + 1))
 
-    zed_notify_pushover "${subject}" "${pathname}"; rv=$?
-    [ "${rv}" -eq 0 ] && num_success=$((num_success + 1))
-    [ "${rv}" -eq 1 ] && num_failure=$((num_failure + 1))
-
     [ "${num_success}" -gt 0 ] && return 0
     [ "${num_failure}" -gt 0 ] && return 1
     return 2
@@ -224,8 +224,6 @@ zed_notify()
 # ZED_EMAIL_OPTS.  This undergoes the following keyword substitutions:
 # - @ADDRESS@ is replaced with the space-delimited recipient email address(es)
 # - @SUBJECT@ is replaced with the notification subject
-#   If @SUBJECT@ was omited here, a "Subject: ..." header will be added to notification
-#
 #
 # Arguments
 #   subject: notification subject
@@ -243,7 +241,7 @@ zed_notify()
 #
 zed_notify_email()
 {
-    local subject="${1:-"ZED notification"}"
+    local subject="$1"
     local pathname="${2:-"/dev/null"}"
 
     : "${ZED_EMAIL_PROG:="mail"}"
@@ -260,30 +258,19 @@ zed_notify_email()
     [ -n "${subject}" ] || return 1
     if [ ! -r "${pathname}" ]; then
         zed_log_err \
-                "${ZED_EMAIL_PROG##*/} cannot read \"${pathname}\""
+                "$(basename "${ZED_EMAIL_PROG}") cannot read \"${pathname}\""
         return 1
     fi
 
-    # construct cmdline options
+    ZED_EMAIL_OPTS="$(echo "${ZED_EMAIL_OPTS}" \
-    ZED_EMAIL_OPTS_PARSED="$(echo "${ZED_EMAIL_OPTS}" \
         | sed   -e "s/@ADDRESS@/${ZED_EMAIL_ADDR}/g" \
                 -e "s/@SUBJECT@/${subject}/g")"
 
-    # pipe message to email prog
+    # shellcheck disable=SC2086
-    # shellcheck disable=SC2086,SC2248
+    eval "${ZED_EMAIL_PROG}" ${ZED_EMAIL_OPTS} < "${pathname}" >/dev/null 2>&1
-    {
-        # no subject passed as option?
-        if [ "${ZED_EMAIL_OPTS%@SUBJECT@*}" = "${ZED_EMAIL_OPTS}" ] ; then
-            # inject subject header
-            printf "Subject: %s\n" "${subject}"
-        fi
-        # output message
-        cat "${pathname}"
-    } |
-    eval ${ZED_EMAIL_PROG} ${ZED_EMAIL_OPTS_PARSED} >/dev/null 2>&1
     rv=$?
     if [ "${rv}" -ne 0 ]; then
-        zed_log_err "${ZED_EMAIL_PROG##*/} exit=${rv}"
+        zed_log_err "$(basename "${ZED_EMAIL_PROG}") exit=${rv}"
         return 1
     fi
     return 0
@@ -380,7 +367,7 @@ zed_notify_pushbullet()
 #
 # Notification via Slack Webhook <https://api.slack.com/incoming-webhooks>.
 # The Webhook URL (ZED_SLACK_WEBHOOK_URL) identifies this client to the
-# Slack channel.
+# Slack channel. 
 #
 # Requires awk, curl, and sed executables to be installed in the standard PATH.
 #
@@ -430,7 +417,7 @@ zed_notify_slack_webhook()
 
     # Construct the JSON message for posting.
     #
-    msg_json="$(printf '{"text": "*%s*\\n%s"}' "${subject}" "${msg_body}" )"
+    msg_json="$(printf '{"text": "*%s*\n%s"}' "${subject}" "${msg_body}" )"
 
     # Send the POST request and check for errors.
     #
@@ -450,84 +437,6 @@ zed_notify_slack_webhook()
     return 0
 }
 
-# zed_notify_pushover (subject, pathname)
-#
-# Send a notification via Pushover <https://pushover.net/>.
-# The access token (ZED_PUSHOVER_TOKEN) identifies this client to the
-# Pushover server. The user token (ZED_PUSHOVER_USER) defines the user or
-# group to which the notification will be sent.
-#
-# Requires curl and sed executables to be installed in the standard PATH.
-#
-# References
-#   https://pushover.net/api
-#
-# Arguments
-#   subject: notification subject
-#   pathname: pathname containing the notification message (OPTIONAL)
-#
-# Globals
-#   ZED_PUSHOVER_TOKEN
-#   ZED_PUSHOVER_USER
-#
-# Return
-#   0: notification sent
-#   1: notification failed
-#   2: not configured
-#
-zed_notify_pushover()
-{
-    local subject="$1"
-    local pathname="${2:-"/dev/null"}"
-    local msg_body
-    local msg_out
-    local msg_err
-    local url="https://api.pushover.net/1/messages.json"
-
-    [ -n "${ZED_PUSHOVER_TOKEN}" ] && [ -n "${ZED_PUSHOVER_USER}" ] || return 2
-
-    if [ ! -r "${pathname}" ]; then
-        zed_log_err "pushover cannot read \"${pathname}\""
-        return 1
-    fi
-
-    zed_check_cmd "curl" "sed" || return 1
-
-    # Read the message body in.
-    #
-    msg_body="$(cat "${pathname}")"
-
-    if [ -z "${msg_body}" ]
-    then
-        msg_body=$subject
-        subject=""
-    fi
-
-    # Send the POST request and check for errors.
-    #
-    msg_out="$( \
-        curl \
-        --form-string "token=${ZED_PUSHOVER_TOKEN}" \
-        --form-string "user=${ZED_PUSHOVER_USER}" \
-        --form-string "message=${msg_body}" \
-        --form-string "title=${subject}" \
-        "${url}" \
-        2>/dev/null \
-        )"; rv=$?
-    if [ "${rv}" -ne 0 ]; then
-        zed_log_err "curl exit=${rv}"
-        return 1
-    fi
-    msg_err="$(echo "${msg_out}" \
-        | sed -n -e 's/.*"errors" *:.*\[\(.*\)\].*/\1/p')"
-    if [ -n "${msg_err}" ]; then
-        zed_log_err "pushover \"${msg_err}"\"
-        return 1
-    fi
-    return 0
-}
-
-
 # zed_rate_limit (tag, [interval])
 #
 # Check whether an event of a given type [tag] has already occurred within the
@@ -602,8 +511,10 @@ zed_guid_to_pool()
 		return
 	fi
 
-	guid="$(printf "%u" "$1")"
+	guid=$(printf "%llu" "$1")
-	$ZPOOL get -H -ovalue,name guid | awk '$1 == '"$guid"' {print $2; exit}'
+	if [ -n "$guid" ] ; then
+		$ZPOOL get -H -ovalue,name guid | awk '$1=='"$guid"' {print $2}'
+	fi
 }
 
 # zed_exit_if_ignoring_this_event

cmd/zed/zed.d/zed.rc

@@ -13,9 +13,9 @@
 # Email address of the zpool administrator for receipt of notifications;
 #   multiple addresses can be specified if they are delimited by whitespace.
 # Email will only be sent if ZED_EMAIL_ADDR is defined.
-# Enabled by default; comment to disable.
+# Disabled by default; uncomment to enable.
 #
-ZED_EMAIL_ADDR="root"
+#ZED_EMAIL_ADDR="root"
 
 ##
 # Name or path of executable responsible for sending notifications via email;
@@ -30,7 +30,6 @@ ZED_EMAIL_ADDR="root"
 # The string @SUBJECT@ will be replaced with the notification subject;
 #   this should be protected with quotes to prevent word-splitting.
 # Email will only be sent if ZED_EMAIL_ADDR is defined.
-# If @SUBJECT@ was omited here, a "Subject: ..." header will be added to notification
 #
 #ZED_EMAIL_OPTS="-s '@SUBJECT@' @ADDRESS@"
 
@@ -83,23 +82,6 @@ ZED_EMAIL_ADDR="root"
 #
 #ZED_SLACK_WEBHOOK_URL=""
 
-##
-# Pushover token.
-# This defines the application from which the notification will be sent.
-#   <https://pushover.net/api#registration>
-# Disabled by default; uncomment to enable.
-# ZED_PUSHOVER_USER, below, must also be configured.
-#
-#ZED_PUSHOVER_TOKEN=""
-
-##
-# Pushover user key.
-# This defines which user or group will receive Pushover notifications.
-#  <https://pushover.net/api#identifiers>
-# Disabled by default; uncomment to enable.
-# ZED_PUSHOVER_TOKEN, above, must also be configured.
-#ZED_PUSHOVER_USER=""
-
 ##
 # Default directory for zed state files.
 #
@@ -107,8 +89,8 @@ ZED_EMAIL_ADDR="root"
 
 ##
 # Turn on/off enclosure LEDs when drives get DEGRADED/FAULTED.  This works for
-# device mapper and multipath devices as well.  This works with JBOD enclosures
+# device mapper and multipath devices as well.  Your enclosure must be
-# and NVMe PCI drives (assuming they're supported by Linux in sysfs).
+# supported by the Linux SES driver for this to work.
 #
 ZED_USE_ENCLOSURE_LEDS=1
 

cmd/zed/zed.h

@@ -3,7 +3,7 @@
  *
  * Developed at Lawrence Livermore National Laboratory (LLNL-CODE-403049).
  * Copyright (C) 2013-2014 Lawrence Livermore National Security, LLC.
- * Refer to the OpenZFS git commit log for authoritative copyright attribution.
+ * Refer to the ZoL git commit log for authoritative copyright attribution.
  *
  * The contents of this file are subject to the terms of the
  * Common Development and Distribution License Version 1.0 (CDDL-1.0).
@@ -15,6 +15,11 @@
 #ifndef	ZED_H
 #define	ZED_H
 
+/*
+ * Absolute path for the default zed configuration file.
+ */
+#define	ZED_CONF_FILE		SYSCONFDIR "/zfs/zed.conf"
+
 /*
  * Absolute path for the default zed pid file.
  */
@@ -30,6 +35,16 @@
  */
 #define	ZED_ZEDLET_DIR		SYSCONFDIR "/zfs/zed.d"
 
+/*
+ * Reserved for future use.
+ */
+#define	ZED_MAX_EVENTS		0
+
+/*
+ * Reserved for future use.
+ */
+#define	ZED_MIN_EVENTS		0
+
 /*
  * String prefix for ZED variables passed via environment variables.
  */

cmd/zed/zed_conf.c

@@ -3,7 +3,7 @@
  *
  * Developed at Lawrence Livermore National Laboratory (LLNL-CODE-403049).
  * Copyright (C) 2013-2014 Lawrence Livermore National Security, LLC.
- * Refer to the OpenZFS git commit log for authoritative copyright attribution.
+ * Refer to the ZoL git commit log for authoritative copyright attribution.
  *
  * The contents of this file are subject to the terms of the
  * Common Development and Distribution License Version 1.0 (CDDL-1.0).
@@ -22,7 +22,6 @@
 #include <stdio.h>
 #include <stdlib.h>
 #include <string.h>
-#include <sys/types.h>
 #include <sys/stat.h>
 #include <sys/uio.h>
 #include <unistd.h>
@@ -33,26 +32,43 @@
 #include "zed_strings.h"
 
 /*
- * Initialise the configuration with default values.
+ * Return a new configuration with default values.
  */
-void
+struct zed_conf *
-zed_conf_init(struct zed_conf *zcp)
+zed_conf_create(void)
 {
-	memset(zcp, 0, sizeof (*zcp));
+	struct zed_conf *zcp;
 
-	/* zcp->zfs_hdl opened in zed_event_init() */
+	zcp = calloc(1, sizeof (*zcp));
-	/* zcp->zedlets created in zed_conf_scan_dir() */
+	if (!zcp)
+		goto nomem;
 
-	zcp->pid_fd = -1;		/* opened in zed_conf_write_pid() */
+	zcp->syslog_facility = LOG_DAEMON;
-	zcp->state_fd = -1;		/* opened in zed_conf_open_state() */
+	zcp->min_events = ZED_MIN_EVENTS;
-	zcp->zevent_fd = -1;		/* opened in zed_event_init() */
+	zcp->max_events = ZED_MAX_EVENTS;
+	zcp->pid_fd = -1;
+	zcp->zedlets = NULL;		/* created via zed_conf_scan_dir() */
+	zcp->state_fd = -1;		/* opened via zed_conf_open_state() */
+	zcp->zfs_hdl = NULL;		/* opened via zed_event_init() */
+	zcp->zevent_fd = -1;		/* opened via zed_event_init() */
 
-	zcp->max_jobs = 16;
+	if (!(zcp->conf_file = strdup(ZED_CONF_FILE)))
+		goto nomem;
 
-	if (!(zcp->pid_file = strdup(ZED_PID_FILE)) ||
+	if (!(zcp->pid_file = strdup(ZED_PID_FILE)))
-	    !(zcp->zedlet_dir = strdup(ZED_ZEDLET_DIR)) ||
+		goto nomem;
-	    !(zcp->state_file = strdup(ZED_STATE_FILE)))
+
-		zed_log_die("Failed to create conf: %s", strerror(errno));
+	if (!(zcp->zedlet_dir = strdup(ZED_ZEDLET_DIR)))
+		goto nomem;
+
+	if (!(zcp->state_file = strdup(ZED_STATE_FILE)))
+		goto nomem;
+
+	return (zcp);
+
+nomem:
+	zed_log_die("Failed to create conf: %s", strerror(errno));
+	return (NULL);
 }
 
 /*
@@ -63,6 +79,9 @@ zed_conf_init(struct zed_conf *zcp)
 void
 zed_conf_destroy(struct zed_conf *zcp)
 {
+	if (!zcp)
+		return;
+
 	if (zcp->state_fd >= 0) {
 		if (close(zcp->state_fd) < 0)
 			zed_log_msg(LOG_WARNING,
@@ -83,6 +102,10 @@ zed_conf_destroy(struct zed_conf *zcp)
 			    zcp->pid_file, strerror(errno));
 		zcp->pid_fd = -1;
 	}
+	if (zcp->conf_file) {
+		free(zcp->conf_file);
+		zcp->conf_file = NULL;
+	}
 	if (zcp->pid_file) {
 		free(zcp->pid_file);
 		zcp->pid_file = NULL;
@@ -99,6 +122,7 @@ zed_conf_destroy(struct zed_conf *zcp)
 		zed_strings_destroy(zcp->zedlets);
 		zcp->zedlets = NULL;
 	}
+	free(zcp);
 }
 
 /*
@@ -108,52 +132,46 @@ zed_conf_destroy(struct zed_conf *zcp)
  * otherwise, output to stderr and exit with a failure status.
  */
 static void
-_zed_conf_display_help(const char *prog, boolean_t got_err)
+_zed_conf_display_help(const char *prog, int got_err)
 {
-	struct opt { const char *o, *d, *v; };
-
 	FILE *fp = got_err ? stderr : stdout;
-
+	int w1 = 4;			/* width of leading whitespace */
-	struct opt *oo;
+	int w2 = 8;			/* width of L-justified option field */
-	struct opt iopts[] = {
-		{ .o = "-h", .d = "Display help" },
-		{ .o = "-L", .d = "Display license information" },
-		{ .o = "-V", .d = "Display version information" },
-		{},
-	};
-	struct opt nopts[] = {
-		{ .o = "-v", .d = "Be verbose" },
-		{ .o = "-f", .d = "Force daemon to run" },
-		{ .o = "-F", .d = "Run daemon in the foreground" },
-		{ .o = "-I",
-		    .d = "Idle daemon until kernel module is (re)loaded" },
-		{ .o = "-M", .d = "Lock all pages in memory" },
-		{ .o = "-P", .d = "$PATH for ZED to use (only used by ZTS)" },
-		{ .o = "-Z", .d = "Zero state file" },
-		{},
-	};
-	struct opt vopts[] = {
-		{ .o = "-d DIR", .d = "Read enabled ZEDLETs from DIR.",
-		    .v = ZED_ZEDLET_DIR },
-		{ .o = "-p FILE", .d = "Write daemon's PID to FILE.",
-		    .v = ZED_PID_FILE },
-		{ .o = "-s FILE", .d = "Write daemon's state to FILE.",
-		    .v = ZED_STATE_FILE },
-		{ .o = "-j JOBS", .d = "Start at most JOBS at once.",
-		    .v = "16" },
-		{},
-	};
 
 	fprintf(fp, "Usage: %s [OPTION]...\n", (prog ? prog : "zed"));
 	fprintf(fp, "\n");
-	for (oo = iopts; oo->o; ++oo)
+	fprintf(fp, "%*c%*s %s\n", w1, 0x20, -w2, "-h",
-		fprintf(fp, "    %*s %s\n", -8, oo->o, oo->d);
+	    "Display help.");
+	fprintf(fp, "%*c%*s %s\n", w1, 0x20, -w2, "-L",
+	    "Display license information.");
+	fprintf(fp, "%*c%*s %s\n", w1, 0x20, -w2, "-V",
+	    "Display version information.");
 	fprintf(fp, "\n");
-	for (oo = nopts; oo->o; ++oo)
+	fprintf(fp, "%*c%*s %s\n", w1, 0x20, -w2, "-v",
-		fprintf(fp, "    %*s %s\n", -8, oo->o, oo->d);
+	    "Be verbose.");
+	fprintf(fp, "%*c%*s %s\n", w1, 0x20, -w2, "-f",
+	    "Force daemon to run.");
+	fprintf(fp, "%*c%*s %s\n", w1, 0x20, -w2, "-F",
+	    "Run daemon in the foreground.");
+	fprintf(fp, "%*c%*s %s\n", w1, 0x20, -w2, "-I",
+	    "Idle daemon until kernel module is (re)loaded.");
+	fprintf(fp, "%*c%*s %s\n", w1, 0x20, -w2, "-M",
+	    "Lock all pages in memory.");
+	fprintf(fp, "%*c%*s %s\n", w1, 0x20, -w2, "-P",
+	    "$PATH for ZED to use (only used by ZTS).");
+	fprintf(fp, "%*c%*s %s\n", w1, 0x20, -w2, "-Z",
+	    "Zero state file.");
 	fprintf(fp, "\n");
-	for (oo = vopts; oo->o; ++oo)
+#if 0
-		fprintf(fp, "    %*s %s [%s]\n", -8, oo->o, oo->d, oo->v);
+	fprintf(fp, "%*c%*s %s [%s]\n", w1, 0x20, -w2, "-c FILE",
+	    "Read configuration from FILE.", ZED_CONF_FILE);
+#endif
+	fprintf(fp, "%*c%*s %s [%s]\n", w1, 0x20, -w2, "-d DIR",
+	    "Read enabled ZEDLETs from DIR.", ZED_ZEDLET_DIR);
+	fprintf(fp, "%*c%*s %s [%s]\n", w1, 0x20, -w2, "-p FILE",
+	    "Write daemon's PID to FILE.", ZED_PID_FILE);
+	fprintf(fp, "%*c%*s %s [%s]\n", w1, 0x20, -w2, "-s FILE",
+	    "Write daemon's state to FILE.", ZED_STATE_FILE);
 	fprintf(fp, "\n");
 
 	exit(got_err ? EXIT_FAILURE : EXIT_SUCCESS);
@@ -165,14 +183,20 @@ _zed_conf_display_help(const char *prog, boolean_t got_err)
 static void
 _zed_conf_display_license(void)
 {
-	printf(
+	const char **pp;
-	    "The ZFS Event Daemon (ZED) is distributed under the terms of the\n"
+	const char *text[] = {
-	    "  Common Development and Distribution License (CDDL-1.0)\n"
+	    "The ZFS Event Daemon (ZED) is distributed under the terms of the",
-	    "  <http://opensource.org/licenses/CDDL-1.0>.\n"
+	    "  Common Development and Distribution License (CDDL-1.0)",
-	    "\n"
+	    "  <http://opensource.org/licenses/CDDL-1.0>.",
+	    "",
 	    "Developed at Lawrence Livermore National Laboratory"
-	    " (LLNL-CODE-403049).\n"
+	    " (LLNL-CODE-403049).",
-	    "\n");
+	    "",
+	    NULL
+	};
+
+	for (pp = text; *pp; pp++)
+		printf("%s\n", *pp);
 
 	exit(EXIT_SUCCESS);
 }
@@ -207,19 +231,16 @@ _zed_conf_parse_path(char **resultp, const char *path)
 
 	if (path[0] == '/') {
 		*resultp = strdup(path);
+	} else if (!getcwd(buf, sizeof (buf))) {
+		zed_log_die("Failed to get current working dir: %s",
+		    strerror(errno));
+	} else if (strlcat(buf, "/", sizeof (buf)) >= sizeof (buf)) {
+		zed_log_die("Failed to copy path: %s", strerror(ENAMETOOLONG));
+	} else if (strlcat(buf, path, sizeof (buf)) >= sizeof (buf)) {
+		zed_log_die("Failed to copy path: %s", strerror(ENAMETOOLONG));
 	} else {
-		if (!getcwd(buf, sizeof (buf)))
-			zed_log_die("Failed to get current working dir: %s",
-			    strerror(errno));
-
-		if (strlcat(buf, "/", sizeof (buf)) >= sizeof (buf) ||
-		    strlcat(buf, path, sizeof (buf)) >= sizeof (buf))
-			zed_log_die("Failed to copy path: %s",
-			    strerror(ENAMETOOLONG));
-
 		*resultp = strdup(buf);
 	}
-
 	if (!*resultp)
 		zed_log_die("Failed to copy path: %s", strerror(ENOMEM));
 }
@@ -230,9 +251,8 @@ _zed_conf_parse_path(char **resultp, const char *path)
 void
 zed_conf_parse_opts(struct zed_conf *zcp, int argc, char **argv)
 {
-	const char * const opts = ":hLVd:p:P:s:vfFMZIj:";
+	const char * const opts = ":hLVc:d:p:P:s:vfFMZI";
 	int opt;
-	unsigned long raw;
 
 	if (!zcp || !argv || !argv[0])
 		zed_log_die("Failed to parse options: Internal error");
@@ -242,7 +262,7 @@ zed_conf_parse_opts(struct zed_conf *zcp, int argc, char **argv)
 	while ((opt = getopt(argc, argv, opts)) != -1) {
 		switch (opt) {
 		case 'h':
-			_zed_conf_display_help(argv[0], B_FALSE);
+			_zed_conf_display_help(argv[0], EXIT_SUCCESS);
 			break;
 		case 'L':
 			_zed_conf_display_license();
@@ -250,6 +270,9 @@ zed_conf_parse_opts(struct zed_conf *zcp, int argc, char **argv)
 		case 'V':
 			_zed_conf_display_version();
 			break;
+		case 'c':
+			_zed_conf_parse_path(&zcp->conf_file, optarg);
+			break;
 		case 'd':
 			_zed_conf_parse_path(&zcp->zedlet_dir, optarg);
 			break;
@@ -280,30 +303,31 @@ zed_conf_parse_opts(struct zed_conf *zcp, int argc, char **argv)
 		case 'Z':
 			zcp->do_zero = 1;
 			break;
-		case 'j':
-			errno = 0;
-			raw = strtoul(optarg, NULL, 0);
-			if (errno == ERANGE || raw > INT16_MAX) {
-				zed_log_die("%lu is too many jobs", raw);
-			} if (raw == 0) {
-				zed_log_die("0 jobs makes no sense");
-			} else {
-				zcp->max_jobs = raw;
-			}
-			break;
 		case '?':
 		default:
 			if (optopt == '?')
-				_zed_conf_display_help(argv[0], B_FALSE);
+				_zed_conf_display_help(argv[0], EXIT_SUCCESS);
 
-			fprintf(stderr, "%s: Invalid option '-%c'\n\n",
+			fprintf(stderr, "%s: %s '-%c'\n\n", argv[0],
-			    argv[0], optopt);
+			    "Invalid option", optopt);
-			_zed_conf_display_help(argv[0], B_TRUE);
+			_zed_conf_display_help(argv[0], EXIT_FAILURE);
 			break;
 		}
 	}
 }
 
+/*
+ * Parse the configuration file into the configuration [zcp].
+ *
+ * FIXME: Not yet implemented.
+ */
+void
+zed_conf_parse_file(struct zed_conf *zcp)
+{
+	if (!zcp)
+		zed_log_die("Failed to parse config: %s", strerror(EINVAL));
+}
+
 /*
  * Scan the [zcp] zedlet_dir for files to exec based on the event class.
  * Files must be executable by user, but not writable by group or other.
@@ -311,6 +335,8 @@ zed_conf_parse_opts(struct zed_conf *zcp, int argc, char **argv)
  *
  * Return 0 on success with an updated set of zedlets,
  * or -1 on error with errno set.
+ *
+ * FIXME: Check if zedlet_dir and all parent dirs are secure.
  */
 int
 zed_conf_scan_dir(struct zed_conf *zcp)
@@ -426,6 +452,8 @@ zed_conf_scan_dir(struct zed_conf *zcp)
 int
 zed_conf_write_pid(struct zed_conf *zcp)
 {
+	const mode_t dirmode = S_IRWXU | S_IRGRP | S_IXGRP | S_IROTH | S_IXOTH;
+	const mode_t filemode = S_IRUSR | S_IWUSR | S_IRGRP | S_IROTH;
 	char buf[PATH_MAX];
 	int n;
 	char *p;
@@ -453,7 +481,7 @@ zed_conf_write_pid(struct zed_conf *zcp)
 	if (p)
 		*p = '\0';
 
-	if ((mkdirp(buf, 0755) < 0) && (errno != EEXIST)) {
+	if ((mkdirp(buf, dirmode) < 0) && (errno != EEXIST)) {
 		zed_log_msg(LOG_ERR, "Failed to create directory \"%s\": %s",
 		    buf, strerror(errno));
 		goto err;
@@ -463,7 +491,7 @@ zed_conf_write_pid(struct zed_conf *zcp)
 	 */
 	mask = umask(0);
 	umask(mask | 022);
-	zcp->pid_fd = open(zcp->pid_file, O_RDWR | O_CREAT | O_CLOEXEC, 0644);
+	zcp->pid_fd = open(zcp->pid_file, (O_RDWR | O_CREAT), filemode);
 	umask(mask);
 	if (zcp->pid_fd < 0) {
 		zed_log_msg(LOG_ERR, "Failed to open PID file \"%s\": %s",
@@ -500,7 +528,7 @@ zed_conf_write_pid(struct zed_conf *zcp)
 		errno = ERANGE;
 		zed_log_msg(LOG_ERR, "Failed to write PID file \"%s\": %s",
 		    zcp->pid_file, strerror(errno));
-	} else if (write(zcp->pid_fd, buf, n) != n) {
+	} else if (zed_file_write_n(zcp->pid_fd, buf, n) != n) {
 		zed_log_msg(LOG_ERR, "Failed to write PID file \"%s\": %s",
 		    zcp->pid_file, strerror(errno));
 	} else if (fdatasync(zcp->pid_fd) < 0) {
@@ -528,6 +556,7 @@ int
 zed_conf_open_state(struct zed_conf *zcp)
 {
 	char dirbuf[PATH_MAX];
+	mode_t dirmode = S_IRWXU | S_IRGRP | S_IXGRP | S_IROTH | S_IXOTH;
 	int n;
 	char *p;
 	int rv;
@@ -549,7 +578,7 @@ zed_conf_open_state(struct zed_conf *zcp)
 	if (p)
 		*p = '\0';
 
-	if ((mkdirp(dirbuf, 0755) < 0) && (errno != EEXIST)) {
+	if ((mkdirp(dirbuf, dirmode) < 0) && (errno != EEXIST)) {
 		zed_log_msg(LOG_WARNING,
 		    "Failed to create directory \"%s\": %s",
 		    dirbuf, strerror(errno));
@@ -567,7 +596,7 @@ zed_conf_open_state(struct zed_conf *zcp)
 		(void) unlink(zcp->state_file);
 
 	zcp->state_fd = open(zcp->state_file,
-	    O_RDWR | O_CREAT | O_CLOEXEC, 0644);
+	    (O_RDWR | O_CREAT), (S_IRUSR | S_IWUSR | S_IRGRP | S_IROTH));
 	if (zcp->state_fd < 0) {
 		zed_log_msg(LOG_WARNING, "Failed to open state file \"%s\": %s",
 		    zcp->state_file, strerror(errno));

cmd/zed/zed_conf.h

@@ -3,7 +3,7 @@
  *
  * Developed at Lawrence Livermore National Laboratory (LLNL-CODE-403049).
  * Copyright (C) 2013-2014 Lawrence Livermore National Security, LLC.
- * Refer to the OpenZFS git commit log for authoritative copyright attribution.
+ * Refer to the ZoL git commit log for authoritative copyright attribution.
  *
  * The contents of this file are subject to the terms of the
  * Common Development and Distribution License Version 1.0 (CDDL-1.0).
@@ -20,39 +20,43 @@
 #include "zed_strings.h"
 
 struct zed_conf {
+	unsigned	do_force:1;		/* true if force enabled */
+	unsigned	do_foreground:1;	/* true if run in foreground */
+	unsigned	do_memlock:1;		/* true if locking memory */
+	unsigned	do_verbose:1;		/* true if verbosity enabled */
+	unsigned	do_zero:1;		/* true if zeroing state */
+	unsigned	do_idle:1;		/* true if idle enabled */
+	int		syslog_facility;	/* syslog facility value */
+	int		min_events;		/* RESERVED FOR FUTURE USE */
+	int		max_events;		/* RESERVED FOR FUTURE USE */
+	char		*conf_file;		/* abs path to config file */
 	char		*pid_file;		/* abs path to pid file */
-	char		*zedlet_dir;		/* abs path to zedlet dir */
-	char		*state_file;		/* abs path to state file */
-
-	libzfs_handle_t	*zfs_hdl;		/* handle to libzfs */
-	zed_strings_t	*zedlets;		/* names of enabled zedlets */
-	char		*path;		/* custom $PATH for zedlets to use */
-
 	int		pid_fd;			/* fd to pid file for lock */
+	char		*zedlet_dir;		/* abs path to zedlet dir */
+	zed_strings_t	*zedlets;		/* names of enabled zedlets */
+	char		*state_file;		/* abs path to state file */
 	int		state_fd;		/* fd to state file */
+	libzfs_handle_t	*zfs_hdl;		/* handle to libzfs */
 	int		zevent_fd;		/* fd for access to zevents */
-
+	char		*path;		/* custom $PATH for zedlets to use */
-	int16_t max_jobs;		/* max zedlets to run at one time */
-
-	boolean_t	do_force:1;		/* true if force enabled */
-	boolean_t	do_foreground:1;	/* true if run in foreground */
-	boolean_t	do_memlock:1;		/* true if locking memory */
-	boolean_t	do_verbose:1;		/* true if verbosity enabled */
-	boolean_t	do_zero:1;		/* true if zeroing state */
-	boolean_t	do_idle:1;		/* true if idle enabled */
 };
 
-void zed_conf_init(struct zed_conf *zcp);
+struct zed_conf *zed_conf_create(void);
+
 void zed_conf_destroy(struct zed_conf *zcp);
 
 void zed_conf_parse_opts(struct zed_conf *zcp, int argc, char **argv);
 
+void zed_conf_parse_file(struct zed_conf *zcp);
+
 int zed_conf_scan_dir(struct zed_conf *zcp);
 
 int zed_conf_write_pid(struct zed_conf *zcp);
 
 int zed_conf_open_state(struct zed_conf *zcp);
+
 int zed_conf_read_state(struct zed_conf *zcp, uint64_t *eidp, int64_t etime[]);
+
 int zed_conf_write_state(struct zed_conf *zcp, uint64_t eid, int64_t etime[]);
 
 #endif	/* !ZED_CONF_H */

cmd/zed/zed_disk_event.c

@@ -72,8 +72,6 @@ zed_udev_event(const char *class, const char *subclass, nvlist_t *nvl)
 		zed_log_msg(LOG_INFO, "\t%s: %s", DEV_PATH, strval);
 	if (nvlist_lookup_string(nvl, DEV_IDENTIFIER, &strval) == 0)
 		zed_log_msg(LOG_INFO, "\t%s: %s", DEV_IDENTIFIER, strval);
-	if (nvlist_lookup_boolean(nvl, DEV_IS_PART) == B_TRUE)
-		zed_log_msg(LOG_INFO, "\t%s: B_TRUE", DEV_IS_PART);
 	if (nvlist_lookup_string(nvl, DEV_PHYS_PATH, &strval) == 0)
 		zed_log_msg(LOG_INFO, "\t%s: %s", DEV_PHYS_PATH, strval);
 	if (nvlist_lookup_uint64(nvl, DEV_SIZE, &numval) == 0)
@@ -215,11 +213,6 @@ zed_udev_monitor(void *arg)
 		if (type != NULL && type[0] != '\0' &&
 		    strcmp(type, "disk") == 0 &&
 		    part != NULL && part[0] != '\0') {
-			zed_log_msg(LOG_INFO,
-			    "%s: skip %s since it has a %s partition already",
-			    __func__,
-			    udev_device_get_property_value(dev, "DEVNAME"),
-			    part);
 			/* skip and wait for partition event */
 			udev_device_unref(dev);
 			continue;
@@ -234,11 +227,6 @@ zed_udev_monitor(void *arg)
 			sectors = udev_device_get_sysattr_value(dev, "size");
 		if (sectors != NULL &&
 		    strtoull(sectors, NULL, 10) < MINIMUM_SECTORS) {
-			zed_log_msg(LOG_INFO,
-			    "%s: %s sectors %s < %llu (minimum)",
-			    __func__,
-			    udev_device_get_property_value(dev, "DEVNAME"),
-			    sectors, MINIMUM_SECTORS);
 			udev_device_unref(dev);
 			continue;
 		}
@@ -391,7 +379,6 @@ zed_disk_event_init()
 		return (-1);
 	}
 
-	pthread_setname_np(g_mon_tid, "udev monitor");
 	zed_log_msg(LOG_INFO, "zed_disk_event_init");
 
 	return (0);

cmd/zed/zed_event.c

@@ -3,7 +3,7 @@
  *
  * Developed at Lawrence Livermore National Laboratory (LLNL-CODE-403049).
  * Copyright (C) 2013-2014 Lawrence Livermore National Security, LLC.
- * Refer to the OpenZFS git commit log for authoritative copyright attribution.
+ * Refer to the ZoL git commit log for authoritative copyright attribution.
  *
  * The contents of this file are subject to the terms of the
  * Common Development and Distribution License Version 1.0 (CDDL-1.0).
@@ -15,7 +15,7 @@
 #include <ctype.h>
 #include <errno.h>
 #include <fcntl.h>
-#include <libzfs_core.h>
+#include <libzfs.h>			/* FIXME: Replace with libzfs_core. */
 #include <paths.h>
 #include <stdarg.h>
 #include <stdio.h>
@@ -54,7 +54,7 @@ zed_event_init(struct zed_conf *zcp)
 		zed_log_die("Failed to initialize libzfs");
 	}
 
-	zcp->zevent_fd = open(ZFS_DEV, O_RDWR | O_CLOEXEC);
+	zcp->zevent_fd = open(ZFS_DEV, O_RDWR);
 	if (zcp->zevent_fd < 0) {
 		if (zcp->do_idle)
 			return (-1);
@@ -96,47 +96,6 @@ zed_event_fini(struct zed_conf *zcp)
 		libzfs_fini(zcp->zfs_hdl);
 		zcp->zfs_hdl = NULL;
 	}
-
-	zed_exec_fini();
-}
-
-static void
-_bump_event_queue_length(void)
-{
-	int zzlm = -1, wr;
-	char qlen_buf[12] = {0}; /* parameter is int => max "-2147483647\n" */
-	long int qlen;
-
-	zzlm = open("/sys/module/zfs/parameters/zfs_zevent_len_max", O_RDWR);
-	if (zzlm < 0)
-		goto done;
-
-	if (read(zzlm, qlen_buf, sizeof (qlen_buf)) < 0)
-		goto done;
-	qlen_buf[sizeof (qlen_buf) - 1] = '\0';
-
-	errno = 0;
-	qlen = strtol(qlen_buf, NULL, 10);
-	if (errno == ERANGE)
-		goto done;
-
-	if (qlen <= 0)
-		qlen = 512; /* default zfs_zevent_len_max value */
-	else
-		qlen *= 2;
-
-	if (qlen > INT_MAX)
-		qlen = INT_MAX;
-	wr = snprintf(qlen_buf, sizeof (qlen_buf), "%ld", qlen);
-
-	if (pwrite(zzlm, qlen_buf, wr, 0) < 0)
-		goto done;
-
-	zed_log_msg(LOG_WARNING, "Bumping queue length to %ld", qlen);
-
-done:
-	if (zzlm > -1)
-		(void) close(zzlm);
 }
 
 /*
@@ -177,7 +136,10 @@ zed_event_seek(struct zed_conf *zcp, uint64_t saved_eid, int64_t saved_etime[])
 
 		if (n_dropped > 0) {
 			zed_log_msg(LOG_WARNING, "Missed %d events", n_dropped);
-			_bump_event_queue_length();
+			/*
+			 * FIXME: Increase max size of event nvlist in
+			 *   /sys/module/zfs/parameters/zfs_zevent_len_max ?
+			 */
 		}
 		if (nvlist_lookup_uint64(nvl, "eid", &eid) != 0) {
 			zed_log_msg(LOG_WARNING, "Failed to lookup zevent eid");
@@ -249,7 +211,7 @@ _zed_event_value_is_hex(const char *name)
  *
  * All environment variables in [zsp] should be added through this function.
  */
-static __attribute__((format(printf, 5, 6))) int
+static int
 _zed_event_add_var(uint64_t eid, zed_strings_t *zsp,
     const char *prefix, const char *name, const char *fmt, ...)
 {
@@ -624,6 +586,8 @@ _zed_event_add_string_array(uint64_t eid, zed_strings_t *zsp,
  * Convert the nvpair [nvp] to a string which is added to the environment
  * of the child process.
  * Return 0 on success, -1 on error.
+ *
+ * FIXME: Refactor with cmd/zpool/zpool_main.c:zpool_do_events_nvprint()?
  */
 static void
 _zed_event_add_nvpair(uint64_t eid, zed_strings_t *zsp, nvpair_t *nvp)
@@ -722,11 +686,23 @@ _zed_event_add_nvpair(uint64_t eid, zed_strings_t *zsp, nvpair_t *nvp)
 		_zed_event_add_var(eid, zsp, prefix, name,
 		    "%llu", (u_longlong_t)i64);
 		break;
+	case DATA_TYPE_NVLIST:
+		_zed_event_add_var(eid, zsp, prefix, name,
+		    "%s", "_NOT_IMPLEMENTED_");			/* FIXME */
+		break;
 	case DATA_TYPE_STRING:
 		(void) nvpair_value_string(nvp, &str);
 		_zed_event_add_var(eid, zsp, prefix, name,
 		    "%s", (str ? str : "<NULL>"));
 		break;
+	case DATA_TYPE_BOOLEAN_ARRAY:
+		_zed_event_add_var(eid, zsp, prefix, name,
+		    "%s", "_NOT_IMPLEMENTED_");			/* FIXME */
+		break;
+	case DATA_TYPE_BYTE_ARRAY:
+		_zed_event_add_var(eid, zsp, prefix, name,
+		    "%s", "_NOT_IMPLEMENTED_");			/* FIXME */
+		break;
 	case DATA_TYPE_INT8_ARRAY:
 		_zed_event_add_int8_array(eid, zsp, prefix, nvp);
 		break;
@@ -754,11 +730,9 @@ _zed_event_add_nvpair(uint64_t eid, zed_strings_t *zsp, nvpair_t *nvp)
 	case DATA_TYPE_STRING_ARRAY:
 		_zed_event_add_string_array(eid, zsp, prefix, nvp);
 		break;
-	case DATA_TYPE_NVLIST:
-	case DATA_TYPE_BOOLEAN_ARRAY:
-	case DATA_TYPE_BYTE_ARRAY:
 	case DATA_TYPE_NVLIST_ARRAY:
-		_zed_event_add_var(eid, zsp, prefix, name, "_NOT_IMPLEMENTED_");
+		_zed_event_add_var(eid, zsp, prefix, name,
+		    "%s", "_NOT_IMPLEMENTED_");			/* FIXME */
 		break;
 	default:
 		errno = EINVAL;
@@ -938,7 +912,10 @@ zed_event_service(struct zed_conf *zcp)
 
 	if (n_dropped > 0) {
 		zed_log_msg(LOG_WARNING, "Missed %d events", n_dropped);
-		_bump_event_queue_length();
+		/*
+		 * FIXME: Increase max size of event nvlist in
+		 * /sys/module/zfs/parameters/zfs_zevent_len_max ?
+		 */
 	}
 	if (nvlist_lookup_uint64(nvl, "eid", &eid) != 0) {
 		zed_log_msg(LOG_WARNING, "Failed to lookup zevent eid");
@@ -976,7 +953,8 @@ zed_event_service(struct zed_conf *zcp)
 
 		_zed_event_add_time_strings(eid, zsp, etime);
 
-		zed_exec_process(eid, class, subclass, zcp, zsp);
+		zed_exec_process(eid, class, subclass,
+		    zcp->zedlet_dir, zcp->zedlets, zsp, zcp->zevent_fd);
 
 		zed_conf_write_state(zcp, eid, etime);
 

cmd/zed/zed_event.h

@@ -3,7 +3,7 @@
  *
  * Developed at Lawrence Livermore National Laboratory (LLNL-CODE-403049).
  * Copyright (C) 2013-2014 Lawrence Livermore National Security, LLC.
- * Refer to the OpenZFS git commit log for authoritative copyright attribution.
+ * Refer to the ZoL git commit log for authoritative copyright attribution.
  *
  * The contents of this file are subject to the terms of the
  * Common Development and Distribution License Version 1.0 (CDDL-1.0).

cmd/zed/zed_exec.c

@@ -3,7 +3,7 @@
  *
  * Developed at Lawrence Livermore National Laboratory (LLNL-CODE-403049).
  * Copyright (C) 2013-2014 Lawrence Livermore National Security, LLC.
- * Refer to the OpenZFS git commit log for authoritative copyright attribution.
+ * Refer to the ZoL git commit log for authoritative copyright attribution.
  *
  * The contents of this file are subject to the terms of the
  * Common Development and Distribution License Version 1.0 (CDDL-1.0).
@@ -18,55 +18,17 @@
 #include <fcntl.h>
 #include <stdlib.h>
 #include <string.h>
-#include <stddef.h>
-#include <sys/avl.h>
-#include <sys/resource.h>
 #include <sys/stat.h>
 #include <sys/wait.h>
 #include <time.h>
 #include <unistd.h>
-#include <pthread.h>
-#include <signal.h>
-
 #include "zed_exec.h"
+#include "zed_file.h"
 #include "zed_log.h"
 #include "zed_strings.h"
 
 #define	ZEVENT_FILENO	3
 
-struct launched_process_node {
-	avl_node_t node;
-	pid_t pid;
-	uint64_t eid;
-	char *name;
-};
-
-static int
-_launched_process_node_compare(const void *x1, const void *x2)
-{
-	pid_t p1;
-	pid_t p2;
-
-	assert(x1 != NULL);
-	assert(x2 != NULL);
-
-	p1 = ((const struct launched_process_node *) x1)->pid;
-	p2 = ((const struct launched_process_node *) x2)->pid;
-
-	if (p1 < p2)
-		return (-1);
-	else if (p1 == p2)
-		return (0);
-	else
-		return (1);
-}
-
-static pthread_t _reap_children_tid = (pthread_t)-1;
-static volatile boolean_t _reap_children_stop;
-static avl_tree_t _launched_processes;
-static pthread_mutex_t _launched_processes_lock = PTHREAD_MUTEX_INITIALIZER;
-static int16_t _launched_processes_limit;
-
 /*
  * Create an environment string array for passing to execve() using the
  * NAME=VALUE strings in container [zsp].
@@ -117,26 +79,20 @@ _zed_exec_create_env(zed_strings_t *zsp)
  */
 static void
 _zed_exec_fork_child(uint64_t eid, const char *dir, const char *prog,
-    char *env[], int zfd, boolean_t in_foreground)
+    char *env[], int zfd)
 {
 	char path[PATH_MAX];
 	int n;
 	pid_t pid;
 	int fd;
-	struct launched_process_node *node;
+	pid_t wpid;
-	sigset_t mask;
+	int status;
-	struct timespec launch_timeout =
-		{ .tv_sec = 0, .tv_nsec = 200 * 1000 * 1000, };
 
 	assert(dir != NULL);
 	assert(prog != NULL);
 	assert(env != NULL);
 	assert(zfd >= 0);
 
-	while (__atomic_load_n(&_launched_processes_limit,
-	    __ATOMIC_SEQ_CST) <= 0)
-		(void) nanosleep(&launch_timeout, NULL);
-
 	n = snprintf(path, sizeof (path), "%s/%s", dir, prog);
 	if ((n < 0) || (n >= sizeof (path))) {
 		zed_log_msg(LOG_WARNING,
@@ -144,179 +100,101 @@ _zed_exec_fork_child(uint64_t eid, const char *dir, const char *prog,
 		    prog, eid, strerror(ENAMETOOLONG));
 		return;
 	}
-	(void) pthread_mutex_lock(&_launched_processes_lock);
 	pid = fork();
 	if (pid < 0) {
-		(void) pthread_mutex_unlock(&_launched_processes_lock);
 		zed_log_msg(LOG_WARNING,
 		    "Failed to fork \"%s\" for eid=%llu: %s",
 		    prog, eid, strerror(errno));
 		return;
 	} else if (pid == 0) {
-		(void) sigemptyset(&mask);
-		(void) sigprocmask(SIG_SETMASK, &mask, NULL);
-
 		(void) umask(022);
-		if (in_foreground && /* we're already devnulled if daemonised */
+		if ((fd = open("/dev/null", O_RDWR)) != -1) {
-		    (fd = open("/dev/null", O_RDWR | O_CLOEXEC)) != -1) {
 			(void) dup2(fd, STDIN_FILENO);
 			(void) dup2(fd, STDOUT_FILENO);
 			(void) dup2(fd, STDERR_FILENO);
 		}
 		(void) dup2(zfd, ZEVENT_FILENO);
+		zed_file_close_from(ZEVENT_FILENO + 1);
 		execle(path, prog, NULL, env);
 		_exit(127);
 	}
 
 	/* parent process */
 
-	node = calloc(1, sizeof (*node));
-	if (node) {
-		node->pid = pid;
-		node->eid = eid;
-		node->name = strdup(prog);
-
-		avl_add(&_launched_processes, node);
-	}
-	(void) pthread_mutex_unlock(&_launched_processes_lock);
-
-	__atomic_sub_fetch(&_launched_processes_limit, 1, __ATOMIC_SEQ_CST);
 	zed_log_msg(LOG_INFO, "Invoking \"%s\" eid=%llu pid=%d",
 	    prog, eid, pid);
-}
 
-static void
+	/* FIXME: Timeout rogue child processes with sigalarm? */
-_nop(int sig)
-{}
 
-static void *
+	/*
-_reap_children(void *arg)
+	 * Wait for child process using WNOHANG to limit
-{
+	 * the time spent waiting to 10 seconds (10,000ms).
-	struct launched_process_node node, *pnode;
+	 */
-	pid_t pid;
+	for (n = 0; n < 1000; n++) {
-	int status;
+		wpid = waitpid(pid, &status, WNOHANG);
-	struct rusage usage;
+		if (wpid == (pid_t)-1) {
-	struct sigaction sa = {};
+			if (errno == EINTR)
+				continue;
+			zed_log_msg(LOG_WARNING,
+			    "Failed to wait for \"%s\" eid=%llu pid=%d",
+			    prog, eid, pid);
+			break;
+		} else if (wpid == 0) {
+			struct timespec t;
 
-	(void) sigfillset(&sa.sa_mask);
+			/* child still running */
-	(void) sigdelset(&sa.sa_mask, SIGCHLD);
+			t.tv_sec = 0;
-	(void) pthread_sigmask(SIG_SETMASK, &sa.sa_mask, NULL);
+			t.tv_nsec = 10000000;	/* 10ms */
-
+			(void) nanosleep(&t, NULL);
-	(void) sigemptyset(&sa.sa_mask);
+			continue;
-	sa.sa_handler = _nop;
-	sa.sa_flags = SA_NOCLDSTOP;
-	(void) sigaction(SIGCHLD, &sa, NULL);
-
-	for (_reap_children_stop = B_FALSE; !_reap_children_stop; ) {
-		(void) pthread_mutex_lock(&_launched_processes_lock);
-		pid = wait4(0, &status, WNOHANG, &usage);
-
-		if (pid == 0 || pid == (pid_t)-1) {
-			(void) pthread_mutex_unlock(&_launched_processes_lock);
-			if (pid == 0 || errno == ECHILD)
-				pause();
-			else if (errno != EINTR)
-				zed_log_msg(LOG_WARNING,
-				    "Failed to wait for children: %s",
-				    strerror(errno));
-		} else {
-			memset(&node, 0, sizeof (node));
-			node.pid = pid;
-			pnode = avl_find(&_launched_processes, &node, NULL);
-			if (pnode) {
-				memcpy(&node, pnode, sizeof (node));
-
-				avl_remove(&_launched_processes, pnode);
-				free(pnode);
-			}
-			(void) pthread_mutex_unlock(&_launched_processes_lock);
-			__atomic_add_fetch(&_launched_processes_limit, 1,
-			    __ATOMIC_SEQ_CST);
-
-			usage.ru_utime.tv_sec += usage.ru_stime.tv_sec;
-			usage.ru_utime.tv_usec += usage.ru_stime.tv_usec;
-			usage.ru_utime.tv_sec +=
-			    usage.ru_utime.tv_usec / (1000 * 1000);
-			usage.ru_utime.tv_usec %= 1000 * 1000;
-
-			if (WIFEXITED(status)) {
-				zed_log_msg(LOG_INFO,
-				    "Finished \"%s\" eid=%llu pid=%d "
-				    "time=%llu.%06us exit=%d",
-				    node.name, node.eid, pid,
-				    (unsigned long long) usage.ru_utime.tv_sec,
-				    (unsigned int) usage.ru_utime.tv_usec,
-				    WEXITSTATUS(status));
-			} else if (WIFSIGNALED(status)) {
-				zed_log_msg(LOG_INFO,
-				    "Finished \"%s\" eid=%llu pid=%d "
-				    "time=%llu.%06us sig=%d/%s",
-				    node.name, node.eid, pid,
-				    (unsigned long long) usage.ru_utime.tv_sec,
-				    (unsigned int) usage.ru_utime.tv_usec,
-				    WTERMSIG(status),
-				    strsignal(WTERMSIG(status)));
-			} else {
-				zed_log_msg(LOG_INFO,
-				    "Finished \"%s\" eid=%llu pid=%d "
-				    "time=%llu.%06us status=0x%X",
-				    node.name, node.eid,
-				    (unsigned long long) usage.ru_utime.tv_sec,
-				    (unsigned int) usage.ru_utime.tv_usec,
-				    (unsigned int) status);
-			}
-
-			free(node.name);
 		}
+
+		if (WIFEXITED(status)) {
+			zed_log_msg(LOG_INFO,
+			    "Finished \"%s\" eid=%llu pid=%d exit=%d",
+			    prog, eid, pid, WEXITSTATUS(status));
+		} else if (WIFSIGNALED(status)) {
+			zed_log_msg(LOG_INFO,
+			    "Finished \"%s\" eid=%llu pid=%d sig=%d/%s",
+			    prog, eid, pid, WTERMSIG(status),
+			    strsignal(WTERMSIG(status)));
+		} else {
+			zed_log_msg(LOG_INFO,
+			    "Finished \"%s\" eid=%llu pid=%d status=0x%X",
+			    prog, eid, (unsigned int) status);
+		}
+		break;
 	}
 
-	return (NULL);
+	/*
-}
+	 * kill child process after 10 seconds
-
+	 */
-void
+	if (wpid == 0) {
-zed_exec_fini(void)
+		zed_log_msg(LOG_WARNING, "Killing hung \"%s\" pid=%d",
-{
+		    prog, pid);
-	struct launched_process_node *node;
+		(void) kill(pid, SIGKILL);
-	void *ck = NULL;
+		(void) waitpid(pid, &status, 0);
-
-	if (_reap_children_tid == (pthread_t)-1)
-		return;
-
-	_reap_children_stop = B_TRUE;
-	(void) pthread_kill(_reap_children_tid, SIGCHLD);
-	(void) pthread_join(_reap_children_tid, NULL);
-
-	while ((node = avl_destroy_nodes(&_launched_processes, &ck)) != NULL) {
-		free(node->name);
-		free(node);
 	}
-	avl_destroy(&_launched_processes);
-
-	(void) pthread_mutex_destroy(&_launched_processes_lock);
-	(void) pthread_mutex_init(&_launched_processes_lock, NULL);
-
-	_reap_children_tid = (pthread_t)-1;
 }
 
 /*
  * Process the event [eid] by synchronously invoking all zedlets with a
  * matching class prefix.
  *
- * Each executable in [zcp->zedlets] from the directory [zcp->zedlet_dir]
+ * Each executable in [zedlets] from the directory [dir] is matched against
- * is matched against the event's [class], [subclass], and the "all" class
+ * the event's [class], [subclass], and the "all" class (which matches
- * (which matches all events).
+ * all events).  Every zedlet with a matching class prefix is invoked.
- * Every zedlet with a matching class prefix is invoked.
  * The NAME=VALUE strings in [envs] will be passed to the zedlet as
  * environment variables.
  *
- * The file descriptor [zcp->zevent_fd] is the zevent_fd used to track the
+ * The file descriptor [zfd] is the zevent_fd used to track the
  * current cursor location within the zevent nvlist.
  *
  * Return 0 on success, -1 on error.
  */
 int
 zed_exec_process(uint64_t eid, const char *class, const char *subclass,
-    struct zed_conf *zcp, zed_strings_t *envs)
+    const char *dir, zed_strings_t *zedlets, zed_strings_t *envs, int zfd)
 {
 	const char *class_strings[4];
 	const char *allclass = "all";
@@ -325,22 +203,9 @@ zed_exec_process(uint64_t eid, const char *class, const char *subclass,
 	char **e;
 	int n;
 
-	if (!zcp->zedlet_dir || !zcp->zedlets || !envs || zcp->zevent_fd < 0)
+	if (!dir || !zedlets || !envs || zfd < 0)
 		return (-1);
 
-	if (_reap_children_tid == (pthread_t)-1) {
-		_launched_processes_limit = zcp->max_jobs;
-
-		if (pthread_create(&_reap_children_tid, NULL,
-		    _reap_children, NULL) != 0)
-			return (-1);
-		pthread_setname_np(_reap_children_tid, "reap ZEDLETs");
-
-		avl_create(&_launched_processes, _launched_process_node_compare,
-		    sizeof (struct launched_process_node),
-		    offsetof(struct launched_process_node, node));
-	}
-
 	csp = class_strings;
 
 	if (class)
@@ -356,13 +221,11 @@ zed_exec_process(uint64_t eid, const char *class, const char *subclass,
 
 	e = _zed_exec_create_env(envs);
 
-	for (z = zed_strings_first(zcp->zedlets); z;
+	for (z = zed_strings_first(zedlets); z; z = zed_strings_next(zedlets)) {
-	    z = zed_strings_next(zcp->zedlets)) {
 		for (csp = class_strings; *csp; csp++) {
 			n = strlen(*csp);
 			if ((strncmp(z, *csp, n) == 0) && !isalpha(z[n]))
-				_zed_exec_fork_child(eid, zcp->zedlet_dir,
+				_zed_exec_fork_child(eid, dir, z, e, zfd);
-				    z, e, zcp->zevent_fd, zcp->do_foreground);
 		}
 	}
 	free(e);

cmd/zed/zed_exec.h

@@ -3,7 +3,7 @@
  *
  * Developed at Lawrence Livermore National Laboratory (LLNL-CODE-403049).
  * Copyright (C) 2013-2014 Lawrence Livermore National Security, LLC.
- * Refer to the OpenZFS git commit log for authoritative copyright attribution.
+ * Refer to the ZoL git commit log for authoritative copyright attribution.
  *
  * The contents of this file are subject to the terms of the
  * Common Development and Distribution License Version 1.0 (CDDL-1.0).
@@ -17,11 +17,9 @@
 
 #include <stdint.h>
 #include "zed_strings.h"
-#include "zed_conf.h"
-
-void zed_exec_fini(void);
 
 int zed_exec_process(uint64_t eid, const char *class, const char *subclass,
-    struct zed_conf *zcp, zed_strings_t *envs);
+    const char *dir, zed_strings_t *zedlets, zed_strings_t *envs,
+    int zevent_fd);
 
 #endif	/* !ZED_EXEC_H */

cmd/zed/zed_file.c

@@ -3,7 +3,7 @@
  *
  * Developed at Lawrence Livermore National Laboratory (LLNL-CODE-403049).
  * Copyright (C) 2013-2014 Lawrence Livermore National Security, LLC.
- * Refer to the OpenZFS git commit log for authoritative copyright attribution.
+ * Refer to the ZoL git commit log for authoritative copyright attribution.
  *
  * The contents of this file are subject to the terms of the
  * Common Development and Distribution License Version 1.0 (CDDL-1.0).
@@ -12,17 +12,73 @@
  * You may not use this file except in compliance with the license.
  */
 
-#include <dirent.h>
 #include <errno.h>
 #include <fcntl.h>
 #include <limits.h>
 #include <string.h>
+#include <sys/resource.h>
 #include <sys/stat.h>
 #include <sys/types.h>
 #include <unistd.h>
 #include "zed_file.h"
 #include "zed_log.h"
 
+/*
+ * Read up to [n] bytes from [fd] into [buf].
+ * Return the number of bytes read, 0 on EOF, or -1 on error.
+ */
+ssize_t
+zed_file_read_n(int fd, void *buf, size_t n)
+{
+	unsigned char *p;
+	size_t n_left;
+	ssize_t n_read;
+
+	p = buf;
+	n_left = n;
+	while (n_left > 0) {
+		if ((n_read = read(fd, p, n_left)) < 0) {
+			if (errno == EINTR)
+				continue;
+			else
+				return (-1);
+
+		} else if (n_read == 0) {
+			break;
+		}
+		n_left -= n_read;
+		p += n_read;
+	}
+	return (n - n_left);
+}
+
+/*
+ * Write [n] bytes from [buf] out to [fd].
+ * Return the number of bytes written, or -1 on error.
+ */
+ssize_t
+zed_file_write_n(int fd, void *buf, size_t n)
+{
+	const unsigned char *p;
+	size_t n_left;
+	ssize_t n_written;
+
+	p = buf;
+	n_left = n;
+	while (n_left > 0) {
+		if ((n_written = write(fd, p, n_left)) < 0) {
+			if (errno == EINTR)
+				continue;
+			else
+				return (-1);
+
+		}
+		n_left -= n_written;
+		p += n_written;
+	}
+	return (n);
+}
+
 /*
  * Set an exclusive advisory lock on the open file descriptor [fd].
  * Return 0 on success, 1 if a conflicting lock is held by another process,
@@ -104,13 +160,6 @@ zed_file_is_locked(int fd)
 	return (lock.l_pid);
 }
 
-
-#if __APPLE__
-#define	PROC_SELF_FD "/dev/fd"
-#else /* Linux-compatible layout */
-#define	PROC_SELF_FD "/proc/self/fd"
-#endif
-
 /*
  * Close all open file descriptors greater than or equal to [lowfd].
  * Any errors encountered while closing file descriptors are ignored.
@@ -118,24 +167,51 @@ zed_file_is_locked(int fd)
 void
 zed_file_close_from(int lowfd)
 {
-	int errno_bak = errno;
+	const int maxfd_def = 256;
-	int maxfd = 0;
+	int errno_bak;
+	struct rlimit rl;
+	int maxfd;
 	int fd;
-	DIR *fddir;
-	struct dirent *fdent;
 
-	if ((fddir = opendir(PROC_SELF_FD)) != NULL) {
+	errno_bak = errno;
-		while ((fdent = readdir(fddir)) != NULL) {
+
-			fd = atoi(fdent->d_name);
+	if (getrlimit(RLIMIT_NOFILE, &rl) < 0) {
-			if (fd > maxfd && fd != dirfd(fddir))
+		maxfd = maxfd_def;
-				maxfd = fd;
+	} else if (rl.rlim_max == RLIM_INFINITY) {
-		}
+		maxfd = maxfd_def;
-		(void) closedir(fddir);
 	} else {
-		maxfd = sysconf(_SC_OPEN_MAX);
+		maxfd = rl.rlim_max;
 	}
 	for (fd = lowfd; fd < maxfd; fd++)
 		(void) close(fd);
 
 	errno = errno_bak;
 }
+
+/*
+ * Set the CLOEXEC flag on file descriptor [fd] so it will be automatically
+ * closed upon successful execution of one of the exec functions.
+ * Return 0 on success, or -1 on error.
+ *
+ * FIXME: No longer needed?
+ */
+int
+zed_file_close_on_exec(int fd)
+{
+	int flags;
+
+	if (fd < 0) {
+		errno = EBADF;
+		return (-1);
+	}
+	flags = fcntl(fd, F_GETFD);
+	if (flags == -1)
+		return (-1);
+
+	flags |= FD_CLOEXEC;
+
+	if (fcntl(fd, F_SETFD, flags) == -1)
+		return (-1);
+
+	return (0);
+}

cmd/zed/zed_file.h

@@ -3,7 +3,7 @@
  *
  * Developed at Lawrence Livermore National Laboratory (LLNL-CODE-403049).
  * Copyright (C) 2013-2014 Lawrence Livermore National Security, LLC.
- * Refer to the OpenZFS git commit log for authoritative copyright attribution.
+ * Refer to the ZoL git commit log for authoritative copyright attribution.
  *
  * The contents of this file are subject to the terms of the
  * Common Development and Distribution License Version 1.0 (CDDL-1.0).
@@ -18,6 +18,10 @@
 #include <sys/types.h>
 #include <unistd.h>
 
+ssize_t zed_file_read_n(int fd, void *buf, size_t n);
+
+ssize_t zed_file_write_n(int fd, void *buf, size_t n);
+
 int zed_file_lock(int fd);
 
 int zed_file_unlock(int fd);
@@ -26,4 +30,6 @@ pid_t zed_file_is_locked(int fd);
 
 void zed_file_close_from(int fd);
 
+int zed_file_close_on_exec(int fd);
+
 #endif	/* !ZED_FILE_H */

cmd/zed/zed_log.c

@@ -3,7 +3,7 @@
  *
  * Developed at Lawrence Livermore National Laboratory (LLNL-CODE-403049).
  * Copyright (C) 2013-2014 Lawrence Livermore National Security, LLC.
- * Refer to the OpenZFS git commit log for authoritative copyright attribution.
+ * Refer to the ZoL git commit log for authoritative copyright attribution.
  *
  * The contents of this file are subject to the terms of the
  * Common Development and Distribution License Version 1.0 (CDDL-1.0).

cmd/zed/zed_log.h

@@ -3,7 +3,7 @@
  *
  * Developed at Lawrence Livermore National Laboratory (LLNL-CODE-403049).
  * Copyright (C) 2013-2014 Lawrence Livermore National Security, LLC.
- * Refer to the OpenZFS git commit log for authoritative copyright attribution.
+ * Refer to the ZoL git commit log for authoritative copyright attribution.
  *
  * The contents of this file are subject to the terms of the
  * Common Development and Distribution License Version 1.0 (CDDL-1.0).

cmd/zed/zed_strings.c

@@ -3,7 +3,7 @@
  *
  * Developed at Lawrence Livermore National Laboratory (LLNL-CODE-403049).
  * Copyright (C) 2013-2014 Lawrence Livermore National Security, LLC.
- * Refer to the OpenZFS git commit log for authoritative copyright attribution.
+ * Refer to the ZoL git commit log for authoritative copyright attribution.
  *
  * The contents of this file are subject to the terms of the
  * Common Development and Distribution License Version 1.0 (CDDL-1.0).

cmd/zed/zed_strings.h

@@ -3,7 +3,7 @@
  *
  * Developed at Lawrence Livermore National Laboratory (LLNL-CODE-403049).
  * Copyright (C) 2013-2014 Lawrence Livermore National Security, LLC.
- * Refer to the OpenZFS git commit log for authoritative copyright attribution.
+ * Refer to the ZoL git commit log for authoritative copyright attribution.
  *
  * The contents of this file are subject to the terms of the
  * Common Development and Distribution License Version 1.0 (CDDL-1.0).

cmd/zfs/zfs_main.c

@@ -53,6 +53,7 @@
 #include <grp.h>
 #include <pwd.h>
 #include <signal.h>
+#include <sys/debug.h>
 #include <sys/list.h>
 #include <sys/mkdev.h>
 #include <sys/mntent.h>
@@ -70,6 +71,7 @@
 #include <zfs_prop.h>
 #include <zfs_deleg.h>
 #include <libzutil.h>
+#include <libuutil.h>
 #ifdef HAVE_IDMAP
 #include <aclutils.h>
 #include <directory.h>
@@ -268,7 +270,7 @@ get_usage(zfs_help_t idx)
 		return (gettext("\tclone [-p] [-o property=value] ... "
 		    "<snapshot> <filesystem|volume>\n"));
 	case HELP_CREATE:
-		return (gettext("\tcreate [-Pnpuv] [-o property=value] ... "
+		return (gettext("\tcreate [-Pnpv] [-o property=value] ... "
 		    "<filesystem>\n"
 		    "\tcreate [-Pnpsv] [-b blocksize] [-o property=value] ... "
 		    "-V <size> <volume>\n"));
@@ -317,7 +319,7 @@ get_usage(zfs_help_t idx)
 	case HELP_SEND:
 		return (gettext("\tsend [-DnPpRvLecwhb] [-[i|I] snapshot] "
 		    "<snapshot>\n"
-		    "\tsend [-DnvPLecw] [-i snapshot|bookmark] "
+		    "\tsend [-nvPLecw] [-i snapshot|bookmark] "
 		    "<filesystem|volume|snapshot>\n"
 		    "\tsend [-DnPpvLec] [-i bookmark|snapshot] "
 		    "--redact <bookmark> <snapshot>\n"
@@ -728,32 +730,6 @@ finish_progress(char *done)
 	pt_header = NULL;
 }
 
-/* This function checks if the passed fd refers to /dev/null or /dev/zero */
-#ifdef __linux__
-static boolean_t
-is_dev_nullzero(int fd)
-{
-	struct stat st;
-	fstat(fd, &st);
-	return (major(st.st_rdev) == 1 && (minor(st.st_rdev) == 3 /* null */ ||
-	    minor(st.st_rdev) == 5 /* zero */));
-}
-#endif
-
-static void
-note_dev_error(int err, int fd)
-{
-#ifdef __linux__
-	if (err == EINVAL && is_dev_nullzero(fd)) {
-		(void) fprintf(stderr,
-		    gettext("Error: Writing directly to /dev/{null,zero} files"
-		    " on certain kernels is not currently implemented.\n"
-		    "(As a workaround, "
-		    "try \"zfs send [...] | cat > /dev/null\")\n"));
-	}
-#endif
-}
-
 static int
 zfs_mount_and_share(libzfs_handle_t *hdl, const char *dataset, zfs_type_t type)
 {
@@ -916,107 +892,6 @@ usage:
 	return (-1);
 }
 
-/*
- * Return a default volblocksize for the pool which always uses more than
- * half of the data sectors.  This primarily applies to dRAID which always
- * writes full stripe widths.
- */
-static uint64_t
-default_volblocksize(zpool_handle_t *zhp, nvlist_t *props)
-{
-	uint64_t volblocksize, asize = SPA_MINBLOCKSIZE;
-	nvlist_t *tree, **vdevs;
-	uint_t nvdevs;
-
-	nvlist_t *config = zpool_get_config(zhp, NULL);
-
-	if (nvlist_lookup_nvlist(config, ZPOOL_CONFIG_VDEV_TREE, &tree) != 0 ||
-	    nvlist_lookup_nvlist_array(tree, ZPOOL_CONFIG_CHILDREN,
-	    &vdevs, &nvdevs) != 0) {
-		return (ZVOL_DEFAULT_BLOCKSIZE);
-	}
-
-	for (int i = 0; i < nvdevs; i++) {
-		nvlist_t *nv = vdevs[i];
-		uint64_t ashift, ndata, nparity;
-
-		if (nvlist_lookup_uint64(nv, ZPOOL_CONFIG_ASHIFT, &ashift) != 0)
-			continue;
-
-		if (nvlist_lookup_uint64(nv, ZPOOL_CONFIG_DRAID_NDATA,
-		    &ndata) == 0) {
-			/* dRAID minimum allocation width */
-			asize = MAX(asize, ndata * (1ULL << ashift));
-		} else if (nvlist_lookup_uint64(nv, ZPOOL_CONFIG_NPARITY,
-		    &nparity) == 0) {
-			/* raidz minimum allocation width */
-			if (nparity == 1)
-				asize = MAX(asize, 2 * (1ULL << ashift));
-			else
-				asize = MAX(asize, 4 * (1ULL << ashift));
-		} else {
-			/* mirror or (non-redundant) leaf vdev */
-			asize = MAX(asize, 1ULL << ashift);
-		}
-	}
-
-	/*
-	 * Calculate the target volblocksize such that more than half
-	 * of the asize is used. The following table is for 4k sectors.
-	 *
-	 * n   asize   blksz  used  |   n   asize   blksz  used
-	 * -------------------------+---------------------------------
-	 * 1   4,096   8,192  100%  |   9  36,864  32,768   88%
-	 * 2   8,192   8,192  100%  |  10  40,960  32,768   80%
-	 * 3  12,288   8,192   66%  |  11  45,056  32,768   72%
-	 * 4  16,384  16,384  100%  |  12  49,152  32,768   66%
-	 * 5  20,480  16,384   80%  |  13  53,248  32,768   61%
-	 * 6  24,576  16,384   66%  |  14  57,344  32,768   57%
-	 * 7  28,672  16,384   57%  |  15  61,440  32,768   53%
-	 * 8  32,768  32,768  100%  |  16  65,536  65,636  100%
-	 *
-	 * This is primarily a concern for dRAID which always allocates
-	 * a full stripe width.  For dRAID the default stripe width is
-	 * n=8 in which case the volblocksize is set to 32k. Ignoring
-	 * compression there are no unused sectors.  This same reasoning
-	 * applies to raidz[2,3] so target 4 sectors to minimize waste.
-	 */
-	uint64_t tgt_volblocksize = ZVOL_DEFAULT_BLOCKSIZE;
-	while (tgt_volblocksize * 2 <= asize)
-		tgt_volblocksize *= 2;
-
-	const char *prop = zfs_prop_to_name(ZFS_PROP_VOLBLOCKSIZE);
-	if (nvlist_lookup_uint64(props, prop, &volblocksize) == 0) {
-
-		/* Issue a warning when a non-optimal size is requested. */
-		if (volblocksize < ZVOL_DEFAULT_BLOCKSIZE) {
-			(void) fprintf(stderr, gettext("Warning: "
-			    "volblocksize (%llu) is less than the default "
-			    "minimum block size (%llu).\nTo reduce wasted "
-			    "space a volblocksize of %llu is recommended.\n"),
-			    (u_longlong_t)volblocksize,
-			    (u_longlong_t)ZVOL_DEFAULT_BLOCKSIZE,
-			    (u_longlong_t)tgt_volblocksize);
-		} else if (volblocksize < tgt_volblocksize) {
-			(void) fprintf(stderr, gettext("Warning: "
-			    "volblocksize (%llu) is much less than the "
-			    "minimum allocation\nunit (%llu), which wastes "
-			    "at least %llu%% of space. To reduce wasted "
-			    "space,\nuse a larger volblocksize (%llu is "
-			    "recommended), fewer dRAID data disks\n"
-			    "per group, or smaller sector size (ashift).\n"),
-			    (u_longlong_t)volblocksize, (u_longlong_t)asize,
-			    (u_longlong_t)((100 * (asize - volblocksize)) /
-			    asize), (u_longlong_t)tgt_volblocksize);
-		}
-	} else {
-		volblocksize = tgt_volblocksize;
-		fnvlist_add_uint64(props, prop, volblocksize);
-	}
-
-	return (volblocksize);
-}
-
 /*
  * zfs create [-Pnpv] [-o prop=value] ... fs
  * zfs create [-Pnpsv] [-b blocksize] [-o prop=value] ... -V vol size
@@ -1036,8 +911,6 @@ default_volblocksize(zpool_handle_t *zhp, nvlist_t *props)
  * check of arguments and properties, but does not check for permissions,
  * available space, etc.
  *
- * The '-u' flag prevents the newly created file system from being mounted.
- *
  * The '-v' flag is for verbose output.
  *
  * The '-P' flag is used for parseable output.  It implies '-v'.
@@ -1054,19 +927,17 @@ zfs_do_create(int argc, char **argv)
 	boolean_t bflag = B_FALSE;
 	boolean_t parents = B_FALSE;
 	boolean_t dryrun = B_FALSE;
-	boolean_t nomount = B_FALSE;
 	boolean_t verbose = B_FALSE;
 	boolean_t parseable = B_FALSE;
 	int ret = 1;
 	nvlist_t *props;
 	uint64_t intval;
-	char *strval;
 
 	if (nvlist_alloc(&props, NV_UNIQUE_NAME, 0) != 0)
 		nomem();
 
 	/* check options */
-	while ((c = getopt(argc, argv, ":PV:b:nso:puv")) != -1) {
+	while ((c = getopt(argc, argv, ":PV:b:nso:pv")) != -1) {
 		switch (c) {
 		case 'V':
 			type = ZFS_TYPE_VOLUME;
@@ -1113,9 +984,6 @@ zfs_do_create(int argc, char **argv)
 		case 's':
 			noreserve = B_TRUE;
 			break;
-		case 'u':
-			nomount = B_TRUE;
-			break;
 		case 'v':
 			verbose = B_TRUE;
 			break;
@@ -1135,11 +1003,6 @@ zfs_do_create(int argc, char **argv)
 		    "used when creating a volume\n"));
 		goto badusage;
 	}
-	if (nomount && type != ZFS_TYPE_FILESYSTEM) {
-		(void) fprintf(stderr, gettext("'-u' can only be "
-		    "used when creating a filesystem\n"));
-		goto badusage;
-	}
 
 	argc -= optind;
 	argv += optind;
@@ -1155,7 +1018,7 @@ zfs_do_create(int argc, char **argv)
 		goto badusage;
 	}
 
-	if (dryrun || type == ZFS_TYPE_VOLUME) {
+	if (dryrun || (type == ZFS_TYPE_VOLUME && !noreserve)) {
 		char msg[ZFS_MAX_DATASET_NAME_LEN * 2];
 		char *p;
 
@@ -1177,24 +1040,18 @@ zfs_do_create(int argc, char **argv)
 		}
 	}
 
+	/*
+	 * if volsize is not a multiple of volblocksize, round it up to the
+	 * nearest multiple of the volblocksize
+	 */
 	if (type == ZFS_TYPE_VOLUME) {
-		const char *prop = zfs_prop_to_name(ZFS_PROP_VOLBLOCKSIZE);
+		uint64_t volblocksize;
-		uint64_t volblocksize = default_volblocksize(zpool_handle,
-		    real_props);
 
-		if (volblocksize != ZVOL_DEFAULT_BLOCKSIZE &&
+		if (nvlist_lookup_uint64(props,
-		    nvlist_lookup_string(props, prop, &strval) != 0) {
+		    zfs_prop_to_name(ZFS_PROP_VOLBLOCKSIZE),
-			if (asprintf(&strval, "%llu",
+		    &volblocksize) != 0)
-			    (u_longlong_t)volblocksize) == -1)
+			volblocksize = ZVOL_DEFAULT_BLOCKSIZE;
-				nomem();
-			nvlist_add_string(props, prop, strval);
-			free(strval);
-		}
 
-		/*
-		 * If volsize is not a multiple of volblocksize, round it
-		 * up to the nearest multiple of the volblocksize.
-		 */
 		if (volsize % volblocksize) {
 			volsize = P2ROUNDUP_TYPED(volsize, volblocksize,
 			    uint64_t);
@@ -1207,9 +1064,11 @@ zfs_do_create(int argc, char **argv)
 		}
 	}
 
+
 	if (type == ZFS_TYPE_VOLUME && !noreserve) {
 		uint64_t spa_version;
 		zfs_prop_t resv_prop;
+		char *strval;
 
 		spa_version = zpool_get_prop_int(zpool_handle,
 		    ZPOOL_PROP_VERSION, NULL);
@@ -1300,11 +1159,6 @@ zfs_do_create(int argc, char **argv)
 		log_history = B_FALSE;
 	}
 
-	if (nomount) {
-		ret = 0;
-		goto error;
-	}
-
 	ret = zfs_mount_and_share(g_zfs, argv[0], ZFS_TYPE_DATASET);
 error:
 	nvlist_free(props);
@@ -4402,7 +4256,6 @@ zfs_do_send(int argc, char **argv)
 
 	struct option long_options[] = {
 		{"replicate",	no_argument,		NULL, 'R'},
-		{"skip-missing",	no_argument,		NULL, 's'},
 		{"redact",	required_argument,	NULL, 'd'},
 		{"props",	no_argument,		NULL, 'p'},
 		{"parsable",	no_argument,		NULL, 'P'},
@@ -4421,7 +4274,7 @@ zfs_do_send(int argc, char **argv)
 	};
 
 	/* check options */
-	while ((c = getopt_long(argc, argv, ":i:I:RsDpvnPLeht:cwbd:S",
+	while ((c = getopt_long(argc, argv, ":i:I:RDpvnPLeht:cwbd:S",
 	    long_options, NULL)) != -1) {
 		switch (c) {
 		case 'i':
@@ -4438,9 +4291,6 @@ zfs_do_send(int argc, char **argv)
 		case 'R':
 			flags.replicate = B_TRUE;
 			break;
-		case 's':
-			flags.skipmissing = B_TRUE;
-			break;
 		case 'd':
 			redactbook = optarg;
 			break;
@@ -4598,23 +4448,11 @@ zfs_do_send(int argc, char **argv)
 
 		err = zfs_send_saved(zhp, &flags, STDOUT_FILENO,
 		    resume_token);
-		if (err != 0)
-			note_dev_error(errno, STDOUT_FILENO);
 		zfs_close(zhp);
 		return (err != 0);
 	} else if (resume_token != NULL) {
-		err = zfs_send_resume(g_zfs, &flags, STDOUT_FILENO,
+		return (zfs_send_resume(g_zfs, &flags, STDOUT_FILENO,
-		    resume_token);
+		    resume_token));
-		if (err != 0)
-			note_dev_error(errno, STDOUT_FILENO);
-		return (err);
-	}
-
-	if (flags.skipmissing && !flags.replicate) {
-		(void) fprintf(stderr,
-		    gettext("skip-missing flag can only be used in "
-		    "conjunction with replicate\n"));
-		usage(B_FALSE);
 	}
 
 	/*
@@ -4658,8 +4496,6 @@ zfs_do_send(int argc, char **argv)
 		err = zfs_send_one(zhp, fromname, STDOUT_FILENO, &flags,
 		    redactbook);
 		zfs_close(zhp);
-		if (err != 0)
-			note_dev_error(errno, STDOUT_FILENO);
 		return (err != 0);
 	}
 
@@ -4736,7 +4572,6 @@ zfs_do_send(int argc, char **argv)
 		nvlist_free(dbgnv);
 	}
 	zfs_close(zhp);
-	note_dev_error(errno, STDOUT_FILENO);
 
 	return (err != 0);
 }
@@ -6593,7 +6428,7 @@ zfs_do_holds(int argc, char **argv)
 		/*
 		 *  1. collect holds data, set format options
 		 */
-		ret = zfs_for_each(1, argv + i, flags, types, NULL, NULL, limit,
+		ret = zfs_for_each(argc, argv, flags, types, NULL, NULL, limit,
 		    holds_callback, &cb);
 		if (ret != 0)
 			++errors;
@@ -7475,7 +7310,6 @@ unshare_unmount(int op, int argc, char **argv)
 				if (zfs_prop_get_int(zhp, ZFS_PROP_CANMOUNT) ==
 				    ZFS_CANMOUNT_NOAUTO)
 					continue;
-				break;
 			default:
 				break;
 			}
@@ -7672,7 +7506,7 @@ zfs_do_diff(int argc, char **argv)
 	int c;
 	struct sigaction sa;
 
-	while ((c = getopt(argc, argv, "FHth")) != -1) {
+	while ((c = getopt(argc, argv, "FHt")) != -1) {
 		switch (c) {
 		case 'F':
 			flags |= ZFS_DIFF_CLASSIFY;
@@ -7683,9 +7517,6 @@ zfs_do_diff(int argc, char **argv)
 		case 't':
 			flags |= ZFS_DIFF_TIMESTAMP;
 			break;
-		case 'h':
-			flags |= ZFS_DIFF_NO_MANGLE;
-			break;
 		default:
 			(void) fprintf(stderr,
 			    gettext("invalid option '%c'\n"), optopt);

cmd/zfs_ids_to_path/zfs_ids_to_path.c

@@ -35,7 +35,7 @@ libzfs_handle_t *g_zfs;
 static void
 usage(int err)
 {
-	fprintf(stderr, "Usage: zfs_ids_to_path [-v] <pool> <objset id> "
+	fprintf(stderr, "Usage: [-v] zfs_ids_to_path <pool> <objset id> "
 	    "<object id>\n");
 	exit(err);
 }
@@ -63,11 +63,11 @@ main(int argc, char **argv)
 
 	uint64_t objset, object;
 	if (sscanf(argv[1], "%llu", (u_longlong_t *)&objset) != 1) {
-		(void) fprintf(stderr, "Invalid objset id: %s\n", argv[1]);
+		(void) fprintf(stderr, "Invalid objset id: %s\n", argv[2]);
 		usage(2);
 	}
 	if (sscanf(argv[2], "%llu", (u_longlong_t *)&object) != 1) {
-		(void) fprintf(stderr, "Invalid object id: %s\n", argv[2]);
+		(void) fprintf(stderr, "Invalid object id: %s\n", argv[3]);
 		usage(3);
 	}
 	if ((g_zfs = libzfs_init()) == NULL) {
@@ -76,7 +76,7 @@ main(int argc, char **argv)
 	}
 	zpool_handle_t *pool = zpool_open(g_zfs, argv[0]);
 	if (pool == NULL) {
-		fprintf(stderr, "Could not open pool %s\n", argv[0]);
+		fprintf(stderr, "Could not open pool %s\n", argv[1]);
 		libzfs_fini(g_zfs);
 		return (5);
 	}

cmd/zgenhostid/zgenhostid.c

@@ -36,6 +36,8 @@
 #include <time.h>
 #include <unistd.h>
 
+static void usage(void);
+
 static void
 usage(void)
 {
@@ -58,11 +60,12 @@ int
 main(int argc, char **argv)
 {
 	/* default file path, can be optionally set by user */
-	const char *path = "/etc/hostid";
+	char path[PATH_MAX] = "/etc/hostid";
 	/* holds converted user input or lrand48() generated value */
 	unsigned long input_i = 0;
 
 	int opt;
+	int pathlen;
 	int force_fwrite = 0;
 	while ((opt = getopt_long(argc, argv, "fo:h?", 0, 0)) != -1) {
 		switch (opt) {
@@ -70,7 +73,14 @@ main(int argc, char **argv)
 			force_fwrite = 1;
 			break;
 		case 'o':
-			path = optarg;
+			pathlen = snprintf(path, sizeof (path), "%s", optarg);
+			if (pathlen >= sizeof (path)) {
+				fprintf(stderr, "%s\n", strerror(EOVERFLOW));
+				exit(EXIT_FAILURE);
+			} else if (pathlen < 1) {
+				fprintf(stderr, "%s\n", strerror(EINVAL));
+				exit(EXIT_FAILURE);
+			}
 			break;
 		case 'h':
 		case '?':
@@ -108,7 +118,7 @@ main(int argc, char **argv)
 	if (force_fwrite == 0 && stat(path, &fstat) == 0 &&
 	    S_ISREG(fstat.st_mode)) {
 		fprintf(stderr, "%s: %s\n", path, strerror(EEXIST));
-		exit(EXIT_FAILURE);
+			exit(EXIT_FAILURE);
 	}
 
 	/*
@@ -127,7 +137,7 @@ main(int argc, char **argv)
 	}
 
 	/*
-	 * we need just 4 bytes in native endianness
+	 * we need just 4 bytes in native endianess
 	 * not using sethostid() because it may be missing or just a stub
 	 */
 	uint32_t hostid = input_i;

cmd/zpool/Makefile.am

@@ -1,5 +1,4 @@
 include $(top_srcdir)/config/Rules.am
-include $(top_srcdir)/config/Shellcheck.am
 
 AM_CFLAGS += $(LIBBLKID_CFLAGS) $(LIBUUID_CFLAGS)
 
@@ -26,8 +25,7 @@ zpool_LDADD = \
 	$(abs_top_builddir)/lib/libzfs/libzfs.la \
 	$(abs_top_builddir)/lib/libzfs_core/libzfs_core.la \
 	$(abs_top_builddir)/lib/libnvpair/libnvpair.la \
-	$(abs_top_builddir)/lib/libuutil/libuutil.la \
+	$(abs_top_builddir)/lib/libuutil/libuutil.la
-	$(abs_top_builddir)/lib/libzutil/libzutil.la
 
 zpool_LDADD += $(LTLIBINTL)
 
@@ -41,7 +39,7 @@ include $(top_srcdir)/config/CppCheck.am
 zpoolconfdir = $(sysconfdir)/zfs/zpool.d
 zpoolexecdir = $(zfsexecdir)/zpool.d
 
-EXTRA_DIST = zpool.d/README compatibility.d
+EXTRA_DIST = zpool.d/README
 
 dist_zpoolexec_SCRIPTS = \
 	zpool.d/dm-deps \
@@ -131,52 +129,6 @@ zpoolconfdefaults = \
 	test_progress \
 	test_ended
 
-zpoolcompatdir = $(pkgdatadir)/compatibility.d
-
-dist_zpoolcompat_DATA = \
-	compatibility.d/compat-2018 \
-	compatibility.d/compat-2019 \
-	compatibility.d/compat-2020 \
-	compatibility.d/compat-2021 \
-	compatibility.d/freebsd-11.0 \
-	compatibility.d/freebsd-11.2 \
-	compatibility.d/freebsd-11.3 \
-	compatibility.d/freenas-9.10.2 \
-	compatibility.d/grub2 \
-	compatibility.d/openzfsonosx-1.7.0 \
-	compatibility.d/openzfsonosx-1.8.1 \
-	compatibility.d/openzfsonosx-1.9.3 \
-	compatibility.d/openzfs-2.0-freebsd \
-	compatibility.d/openzfs-2.0-linux \
-	compatibility.d/openzfs-2.1-freebsd \
-	compatibility.d/openzfs-2.1-linux \
-	compatibility.d/zol-0.6.1 \
-	compatibility.d/zol-0.6.4 \
-	compatibility.d/zol-0.6.5 \
-	compatibility.d/zol-0.7 \
-	compatibility.d/zol-0.8
-
-# canonical <- alias symbolic link pairs
-# eg: "2018" is a link to "compat-2018"
-zpoolcompatlinks = \
-	"compat-2018		2018" \
-	"compat-2019		2019" \
-	"compat-2020		2020" \
-	"compat-2021		2021" \
-	"freebsd-11.0		freebsd-11.1" \
-	"freebsd-11.0		freenas-11.0" \
-	"freebsd-11.2		freenas-11.2" \
-	"freebsd-11.3		freebsd-11.4" \
-	"freebsd-11.3		freebsd-12.0" \
-	"freebsd-11.3		freebsd-12.1" \
-	"freebsd-11.3		freebsd-12.2" \
-	"freebsd-11.3		freenas-11.3" \
-	"freenas-11.0		freenas-11.1" \
-	"openzfsonosx-1.9.3	openzfsonosx-1.9.4" \
-	"openzfs-2.0-freebsd	truenas-12.0" \
-	"zol-0.7		ubuntu-18.04" \
-	"zol-0.8		ubuntu-20.04"
-
 install-data-hook:
 	$(MKDIR_P) "$(DESTDIR)$(zpoolconfdir)"
 	for f in $(zpoolconfdefaults); do \
@@ -184,6 +136,3 @@ install-data-hook:
 	       -L "$(DESTDIR)$(zpoolconfdir)/$${f}" || \
 	    ln -s "$(zpoolexecdir)/$${f}" "$(DESTDIR)$(zpoolconfdir)"; \
 	done
-	for l in $(zpoolcompatlinks); do \
-		(cd "$(DESTDIR)$(zpoolcompatdir)"; ln -sf $${l} ); \
-	done

cmd/zpool/compatibility.d/compat-2018

@@ -1,12 +0,0 @@
-# Features supported by all Tier 1 platforms as of 2018
-async_destroy
-bookmarks
-embedded_data
-empty_bpobj
-enabled_txg
-extensible_dataset
-filesystem_limits
-hole_birth
-large_blocks
-lz4_compress
-spacemap_histogram

cmd/zpool/compatibility.d/compat-2019

@@ -1,15 +0,0 @@
-# Features supported by all Tier 1 platforms as of 2019
-async_destroy
-bookmarks
-embedded_data
-empty_bpobj
-enabled_txg
-extensible_dataset
-filesystem_limits
-hole_birth
-large_blocks
-lz4_compress
-multi_vdev_crash_dump
-sha512
-skein
-spacemap_histogram

cmd/zpool/compatibility.d/compat-2020

@@ -1,15 +0,0 @@
-# Features supported by all Tier 1 platforms as of 2020
-async_destroy
-bookmarks
-embedded_data
-empty_bpobj
-enabled_txg
-extensible_dataset
-filesystem_limits
-hole_birth
-large_blocks
-lz4_compress
-multi_vdev_crash_dump
-sha512
-skein
-spacemap_histogram

cmd/zpool/compatibility.d/compat-2021

@@ -1,19 +0,0 @@
-# Features supported by all Tier 1 platforms as of 2021
-async_destroy
-bookmarks
-device_removal
-embedded_data
-empty_bpobj
-enabled_txg
-extensible_dataset
-filesystem_limits
-hole_birth
-large_blocks
-lz4_compress
-multi_vdev_crash_dump
-obsolete_counts
-sha512
-skein
-spacemap_histogram
-spacemap_v2
-zpool_checkpoint

cmd/zpool/compatibility.d/freebsd-11.0

@@ -1,15 +0,0 @@
-# Features supported by FreeBSD 11.0
-async_destroy
-bookmarks
-embedded_data
-empty_bpobj
-enabled_txg
-extensible_dataset
-filesystem_limits
-hole_birth
-large_blocks
-lz4_compress
-multi_vdev_crash_dump
-sha512
-skein
-spacemap_histogram

cmd/zpool/compatibility.d/freebsd-11.2

@@ -1,18 +0,0 @@
-# Features supported by FreeBSD 11.2
-async_destroy
-bookmarks
-device_removal
-embedded_data
-empty_bpobj
-enabled_txg
-extensible_dataset
-filesystem_limits
-hole_birth
-large_blocks
-lz4_compress
-multi_vdev_crash_dump
-obsolete_counts
-sha512
-skein
-spacemap_histogram
-zpool_checkpoint

cmd/zpool/compatibility.d/freebsd-11.3

@@ -1,19 +0,0 @@
-# Features supported by FreeBSD 11.3
-async_destroy
-bookmarks
-device_removal
-embedded_data
-empty_bpobj
-enabled_txg
-extensible_dataset
-filesystem_limits
-hole_birth
-large_blocks
-lz4_compress
-multi_vdev_crash_dump
-obsolete_counts
-sha512
-skein
-spacemap_histogram
-spacemap_v2
-zpool_checkpoint

cmd/zpool/compatibility.d/freenas-9.10.2

@@ -1,13 +0,0 @@
-# Features supported by FreeNAS 9.10.2
-async_destroy
-bookmarks
-embedded_data
-empty_bpobj
-enabled_txg
-extensible_dataset
-filesystem_limits
-hole_birth
-large_blocks
-lz4_compress
-multi_vdev_crash_dump
-spacemap_histogram

cmd/zpool/compatibility.d/grub2

@@ -1,12 +0,0 @@
-# Features which are supported by GRUB2
-async_destroy
-bookmarks
-embedded_data
-empty_bpobj
-enabled_txg
-extensible_dataset
-filesystem_limits
-hole_birth
-large_blocks
-lz4_compress
-spacemap_histogram

cmd/zpool/compatibility.d/openzfs-2.0-freebsd

@@ -1,33 +0,0 @@
-# Features supported by OpenZFS 2.0 on FreeBSD
-allocation_classes
-async_destroy
-bookmark_v2
-bookmark_written
-bookmarks
-device_rebuild
-device_removal
-embedded_data
-empty_bpobj
-enabled_txg
-encryption
-extensible_dataset
-filesystem_limits
-hole_birth
-large_blocks
-large_dnode
-livelist
-log_spacemap
-lz4_compress
-multi_vdev_crash_dump
-obsolete_counts
-project_quota
-redacted_datasets
-redaction_bookmarks
-resilver_defer
-sha512
-skein
-spacemap_histogram
-spacemap_v2
-userobj_accounting
-zpool_checkpoint
-zstd_compress

cmd/zpool/compatibility.d/openzfs-2.0-linux

@@ -1,34 +0,0 @@
-# Features supported by OpenZFS 2.0 on Linux
-allocation_classes
-async_destroy
-bookmark_v2
-bookmark_written
-bookmarks
-device_rebuild
-device_removal
-edonr
-embedded_data
-empty_bpobj
-enabled_txg
-encryption
-extensible_dataset
-filesystem_limits
-hole_birth
-large_blocks
-large_dnode
-livelist
-log_spacemap
-lz4_compress
-multi_vdev_crash_dump
-obsolete_counts
-project_quota
-redacted_datasets
-redaction_bookmarks
-resilver_defer
-sha512
-skein
-spacemap_histogram
-spacemap_v2
-userobj_accounting
-zpool_checkpoint
-zstd_compress

cmd/zpool/compatibility.d/openzfs-2.1-freebsd

@@ -1,34 +0,0 @@
-# Features supported by OpenZFS 2.1 on FreeBSD
-allocation_classes
-async_destroy
-bookmark_v2
-bookmark_written
-bookmarks
-device_rebuild
-device_removal
-draid
-embedded_data
-empty_bpobj
-enabled_txg
-encryption
-extensible_dataset
-filesystem_limits
-hole_birth
-large_blocks
-large_dnode
-livelist
-log_spacemap
-lz4_compress
-multi_vdev_crash_dump
-obsolete_counts
-project_quota
-redacted_datasets
-redaction_bookmarks
-resilver_defer
-sha512
-skein
-spacemap_histogram
-spacemap_v2
-userobj_accounting
-zpool_checkpoint
-zstd_compress

cmd/zpool/compatibility.d/openzfs-2.1-linux

@@ -1,35 +0,0 @@
-# Features supported by OpenZFS 2.1 on Linux
-allocation_classes
-async_destroy
-bookmark_v2
-bookmark_written
-bookmarks
-device_rebuild
-device_removal
-draid
-edonr
-embedded_data
-empty_bpobj
-enabled_txg
-encryption
-extensible_dataset
-filesystem_limits
-hole_birth
-large_blocks
-large_dnode
-livelist
-log_spacemap
-lz4_compress
-multi_vdev_crash_dump
-obsolete_counts
-project_quota
-redacted_datasets
-redaction_bookmarks
-resilver_defer
-sha512
-skein
-spacemap_histogram
-spacemap_v2
-userobj_accounting
-zpool_checkpoint
-zstd_compress

cmd/zpool/compatibility.d/openzfsonosx-1.7.0

@@ -1,16 +0,0 @@
-# Features supported by OpenZFSonOSX 1.7.0
-async_destroy
-bookmarks
-edonr
-embedded_data
-empty_bpobj
-enabled_txg
-extensible_dataset
-filesystem_limits
-hole_birth
-large_blocks
-lz4_compress
-multi_vdev_crash_dump
-sha512
-skein
-spacemap_histogram

cmd/zpool/compatibility.d/openzfsonosx-1.8.1

@@ -1,21 +0,0 @@
-# Features supported by OpenZFSonOSX 1.8.1
-async_destroy
-bookmarks
-device_removal
-edonr
-embedded_data
-empty_bpobj
-enabled_txg
-encryption
-extensible_dataset
-filesystem_limits
-hole_birth
-large_blocks
-lz4_compress
-multi_vdev_crash_dump
-obsolete_counts
-sha512
-skein
-spacemap_histogram
-spacemap_v2
-zpool_checkpoint

cmd/zpool/compatibility.d/openzfsonosx-1.9.3

@@ -1,27 +0,0 @@
-# Features supported by OpenZFSonOSX 1.9.3
-allocation_classes
-async_destroy
-bookmark_v2
-bookmarks
-device_removal
-edonr
-embedded_data
-empty_bpobj
-enabled_txg
-encryption
-extensible_dataset
-filesystem_limits
-hole_birth
-large_blocks
-large_dnode
-lz4_compress
-multi_vdev_crash_dump
-obsolete_counts
-project_quota
-resilver_defer
-sha512
-skein
-spacemap_histogram
-spacemap_v2
-userobj_accounting
-zpool_checkpoint

cmd/zpool/compatibility.d/zol-0.6.1

@@ -1,4 +0,0 @@
-# Features supported by ZFSonLinux v0.6.1
-async_destroy
-empty_bpobj
-lz4_compress

cmd/zpool/compatibility.d/zol-0.6.4

@@ -1,10 +0,0 @@
-# Features supported by ZFSonLinux v0.6.4
-async_destroy
-bookmarks
-embedded_data
-empty_bpobj
-enabled_txg
-extensible_dataset
-hole_birth
-lz4_compress
-spacemap_histogram

cmd/zpool/compatibility.d/zol-0.6.5

@@ -1,12 +0,0 @@
-# Features supported by ZFSonLinux v0.6.5
-async_destroy
-bookmarks
-embedded_data
-empty_bpobj
-enabled_txg
-extensible_dataset
-filesystem_limits
-hole_birth
-large_blocks
-lz4_compress
-spacemap_histogram

cmd/zpool/compatibility.d/zol-0.7

@@ -1,18 +0,0 @@
-# Features supported by ZFSonLinux v0.7
-async_destroy
-bookmarks
-edonr
-embedded_data
-empty_bpobj
-enabled_txg
-extensible_dataset
-filesystem_limits
-hole_birth
-large_blocks
-large_dnode
-lz4_compress
-multi_vdev_crash_dump
-sha512
-skein
-spacemap_histogram
-userobj_accounting

cmd/zpool/compatibility.d/zol-0.8

@@ -1,27 +0,0 @@
-# Features supported by ZFSonLinux v0.8
-allocation_classes
-async_destroy
-bookmark_v2
-bookmarks
-device_removal
-edonr
-embedded_data
-empty_bpobj
-enabled_txg
-encryption
-extensible_dataset
-filesystem_limits
-hole_birth
-large_blocks
-large_dnode
-lz4_compress
-multi_vdev_crash_dump
-obsolete_counts
-project_quota
-resilver_defer
-sha512
-skein
-spacemap_histogram
-spacemap_v2
-userobj_accounting
-zpool_checkpoint

cmd/zpool/os/freebsd/zpool_vdev_os.c

@@ -101,18 +101,3 @@ check_sector_size_database(char *path, int *sector_size)
 {
 	return (0);
 }
-
-void
-after_zpool_upgrade(zpool_handle_t *zhp)
-{
-	char bootfs[ZPOOL_MAXPROPLEN];
-
-	if (zpool_get_prop(zhp, ZPOOL_PROP_BOOTFS, bootfs,
-	    sizeof (bootfs), NULL, B_FALSE) == 0 &&
-	    strcmp(bootfs, "-") != 0) {
-		(void) printf(gettext("Pool '%s' has the bootfs "
-		    "property set, you might need to update\nthe boot "
-		    "code. See gptzfsboot(8) and loader.efi(8) for "
-		    "details.\n"), zpool_get_name(zhp));
-	}
-}

cmd/zpool/os/linux/zpool_vdev_os.c

@@ -79,6 +79,9 @@
 
 #include <scsi/scsi.h>
 #include <scsi/sg.h>
+#include <stdio.h>
+#include <string.h>
+#include <unistd.h>
 #include <sys/efi_partition.h>
 #include <sys/stat.h>
 #include <sys/vtoc.h>
@@ -405,8 +408,3 @@ check_device(const char *path, boolean_t force,
 
 	return (error);
 }
-
-void
-after_zpool_upgrade(zpool_handle_t *zhp)
-{
-}

cmd/zpool/zpool.d/dm-deps

@@ -16,12 +16,14 @@ if [ -L "$dev" ] ; then
 	dev=$(readlink "$dev")
 fi
 
-dev="${dev##*/}"
+dev=$(basename "$dev")
 val=""
 if [ -d "/sys/class/block/$dev/slaves" ] ; then
-	# ls -C: output in columns, no newlines, two spaces (change to one)
+	# ls -C: output in columns, no newlines
-	# shellcheck disable=SC2012
+	val=$(ls -C "/sys/class/block/$dev/slaves")
-	val=$(ls -C "/sys/class/block/$dev/slaves" | tr -s '[:space:]' ' ')
+
+	# ls -C will print two spaces between files; change to one space.
+	val=$(echo "$val" | sed -r 's/[[:blank:]]+/ /g')
 fi
 
 echo "dm-deps=$val"

cmd/zpool/zpool.d/iostat

@@ -9,7 +9,7 @@ iostat:		Show iostat values since boot (summary page).
 iostat-1s:	Do a single 1-second iostat sample and show values.
 iostat-10s:	Do a single 10-second iostat sample and show values."
 
-script="${0##*/}"
+script=$(basename "$0")
 if [ "$1" = "-h" ] ; then
 	echo "$helpstr" | grep "$script:" | tr -s '\t' | cut -f 2-
 	exit
@@ -42,7 +42,7 @@ else
 		${brief:+"-y"} \
 		${interval:+"$interval"} \
 		${interval:+"1"} \
-		"$VDEV_UPATH" | grep -v '^$' | tail -n 2)
+		"$VDEV_UPATH" | awk NF | tail -n 2)
 fi
 
 
@@ -61,7 +61,7 @@ fi
 cols=$(echo "$out" | head -n 1)
 
 # Get the values and tab separate them to make them cut-able.
-vals=$(echo "$out" | tail -n 1 | tr -s '[:space:]' '\t')
+vals=$(echo "$out" | tail -n 1 | sed -r 's/[[:blank:]]+/\t/g')
 
 i=0
 for col in $cols ; do

cmd/zpool/zpool.d/lsblk

@@ -48,7 +48,7 @@ size:	Show the disk capacity.
 vendor:	Show the disk vendor.
 lsblk:	Show the disk size, vendor, and model number."
 
-script="${0##*/}"
+script=$(basename "$0")
 
 if [ "$1" = "-h" ] ; then
         echo "$helpstr" | grep "$script:" | tr -s '\t' | cut -f 2-

cmd/zpool/zpool.d/media

@@ -4,23 +4,19 @@
 #
 
 if [ "$1" = "-h" ] ; then
-	echo "Show whether a vdev is a file, hdd, ssd, or iscsi."
+	echo "Show whether a vdev is a file, hdd, or ssd."
 	exit
 fi
 
 if [ -b "$VDEV_UPATH" ]; then
-	device="${VDEV_UPATH##*/}"
+	device=$(basename "$VDEV_UPATH")
-	read -r val 2>/dev/null < "/sys/block/$device/queue/rotational"
+	val=$(cat "/sys/block/$device/queue/rotational" 2>/dev/null)
-	case "$val" in
+	if [ "$val" = "0" ]; then
-		0) MEDIA="ssd" ;;
+		MEDIA="ssd"
-		1) MEDIA="hdd" ;;
+	fi
-	esac
 
-	vpd_pg83="/sys/block/$device/device/vpd_pg83"
+	if [ "$val" = "1" ]; then
-	if [ -f "$vpd_pg83" ]; then
+		MEDIA="hdd"
-		if grep -q --binary "iqn." "$vpd_pg83"; then
-			MEDIA="iscsi"
-		fi
 	fi
 else
 	if [ -f "$VDEV_UPATH" ]; then

cmd/zpool/zpool.d/ses

@@ -11,7 +11,7 @@ fault_led:	Show value of the disk enclosure slot fault LED.
 locate_led:	Show value of the disk enclosure slot locate LED.
 ses:		Show disk's enc, enc device, slot, and fault/locate LED values."
 
-script="${0##*/}"
+script=$(basename "$0")
 if [ "$1" = "-h" ] ; then
 	echo "$helpstr" | grep "$script:" | tr -s '\t' | cut -f 2-
 	exit
@@ -41,13 +41,7 @@ for i in $scripts ; do
 		val=$(ls "$VDEV_ENC_SYSFS_PATH/../device/scsi_generic" 2>/dev/null)
 		;;
 	fault_led)
-		# JBODs fault LED is called 'fault', NVMe fault LED is called
+		val=$(cat "$VDEV_ENC_SYSFS_PATH/fault" 2>/dev/null)
-		# 'attention'.
-		if [ -f "$VDEV_ENC_SYSFS_PATH/fault" ] ; then
-			val=$(cat "$VDEV_ENC_SYSFS_PATH/fault" 2>/dev/null)
-		elif [ -f "$VDEV_ENC_SYSFS_PATH/attention" ] ; then
-			val=$(cat "$VDEV_ENC_SYSFS_PATH/attention" 2>/dev/null)
-		fi
 		;;
 	locate_led)
 		val=$(cat "$VDEV_ENC_SYSFS_PATH/locate" 2>/dev/null)

cmd/zpool/zpool.d/smart

@@ -53,7 +53,7 @@ get_filename_from_dir()
 	num_files=$(find "$dir" -maxdepth 1 -type f | wc -l)
 	mod=$((pid % num_files))
 	i=0
-	find "$dir" -type f -printf '%f\n' | while read -r file ; do
+	find "$dir" -type f -printf "%f\n" | while read -r file ; do
 		if [ "$mod" = "$i" ] ; then
 			echo "$file"
 			break
@@ -62,14 +62,17 @@ get_filename_from_dir()
 	done
 }
 
-script="${0##*/}"
+script=$(basename "$0")
 
 if [ "$1" = "-h" ] ; then
         echo "$helpstr" | grep "$script:" | tr -s '\t' | cut -f 2-
         exit
 fi
 
-if [ -b "$VDEV_UPATH" ] && PATH="/usr/sbin:$PATH" command -v smartctl > /dev/null || [ -n "$samples" ] ; then
+smartctl_path=$(command -v smartctl)
+
+# shellcheck disable=SC2015
+if [ -b "$VDEV_UPATH" ] && [ -x "$smartctl_path" ] || [ -n "$samples" ] ; then
 	if [ -n "$samples" ] ; then
 		# cat a smartctl output text file instead of running smartctl
 		# on a vdev (only used for developer testing).
@@ -77,7 +80,7 @@ if [ -b "$VDEV_UPATH" ] && PATH="/usr/sbin:$PATH" command -v smartctl > /dev/nul
 		echo "file=$file"
 		raw_out=$(cat "$samples/$file")
 	else
-		raw_out=$(sudo smartctl -a "$VDEV_UPATH")
+		raw_out=$(eval "sudo $smartctl_path -a $VDEV_UPATH")
 	fi
 
 	# What kind of drive are we?  Look for the right line in smartctl:
@@ -228,11 +231,11 @@ esac
 with_vals=$(echo "$out" | grep -E "$scripts")
 if [ -n "$with_vals" ]; then
 	echo "$with_vals"
-	without_vals=$(echo "$scripts" | tr '|' '\n' |
+	without_vals=$(echo "$scripts" | tr "|" "\n" |
 		grep -v -E "$(echo "$with_vals" |
 		awk -F "=" '{print $1}')" | awk '{print $0"="}')
 else
-	without_vals=$(echo "$scripts" | tr '|' '\n' | awk '{print $0"="}')
+	without_vals=$(echo "$scripts" | tr "|" "\n" | awk '{print $0"="}')
 fi
 
 if [ -n "$without_vals" ]; then

cmd/zpool/zpool_iter.c

@@ -264,6 +264,51 @@ for_each_pool(int argc, char **argv, boolean_t unavail,
 	return (ret);
 }
 
+static int
+for_each_vdev_cb(zpool_handle_t *zhp, nvlist_t *nv, pool_vdev_iter_f func,
+    void *data)
+{
+	nvlist_t **child;
+	uint_t c, children;
+	int ret = 0;
+	int i;
+	char *type;
+
+	const char *list[] = {
+	    ZPOOL_CONFIG_SPARES,
+	    ZPOOL_CONFIG_L2CACHE,
+	    ZPOOL_CONFIG_CHILDREN
+	};
+
+	for (i = 0; i < ARRAY_SIZE(list); i++) {
+		if (nvlist_lookup_nvlist_array(nv, list[i], &child,
+		    &children) == 0) {
+			for (c = 0; c < children; c++) {
+				uint64_t ishole = 0;
+
+				(void) nvlist_lookup_uint64(child[c],
+				    ZPOOL_CONFIG_IS_HOLE, &ishole);
+
+				if (ishole)
+					continue;
+
+				ret |= for_each_vdev_cb(zhp, child[c], func,
+				    data);
+			}
+		}
+	}
+
+	if (nvlist_lookup_string(nv, ZPOOL_CONFIG_TYPE, &type) != 0)
+		return (ret);
+
+	/* Don't run our function on root vdevs */
+	if (strcmp(type, VDEV_TYPE_ROOT) != 0) {
+		ret |= func(zhp, nv, data);
+	}
+
+	return (ret);
+}
+
 /*
  * This is the equivalent of for_each_pool() for vdevs.  It iterates thorough
  * all vdevs in the pool, ignoring root vdevs and holes, calling func() on
@@ -282,7 +327,7 @@ for_each_vdev(zpool_handle_t *zhp, pool_vdev_iter_f func, void *data)
 		verify(nvlist_lookup_nvlist(config, ZPOOL_CONFIG_VDEV_TREE,
 		    &nvroot) == 0);
 	}
-	return (for_each_vdev_cb((void *) zhp, nvroot, func, data));
+	return (for_each_vdev_cb(zhp, nvroot, func, data));
 }
 
 /*
@@ -449,25 +494,19 @@ vdev_run_cmd(vdev_cmd_data_t *data, char *cmd)
 	/* Setup our custom environment variables */
 	rc = asprintf(&env[1], "VDEV_PATH=%s",
 	    data->path ? data->path : "");
-	if (rc == -1) {
+	if (rc == -1)
-		env[1] = NULL;
 		goto out;
-	}
 
 	rc = asprintf(&env[2], "VDEV_UPATH=%s",
 	    data->upath ? data->upath : "");
-	if (rc == -1) {
+	if (rc == -1)
-		env[2] = NULL;
 		goto out;
-	}
 
 	rc = asprintf(&env[3], "VDEV_ENC_SYSFS_PATH=%s",
 	    data->vdev_enc_sysfs_path ?
 	    data->vdev_enc_sysfs_path : "");
-	if (rc == -1) {
+	if (rc == -1)
-		env[3] = NULL;
 		goto out;
-	}
 
 	/* Run the command */
 	rc = libzfs_run_process_get_stdout_nopath(cmd, argv, env, &lines,
@@ -486,7 +525,8 @@ out:
 
 	/* Start with i = 1 since env[0] was statically allocated */
 	for (i = 1; i < ARRAY_SIZE(env); i++)
-		free(env[i]);
+		if (env[i] != NULL)
+			free(env[i]);
 }
 
 /*
@@ -558,7 +598,7 @@ vdev_run_cmd_thread(void *cb_cmd_data)
 
 /* For each vdev in the pool run a command */
 static int
-for_each_vdev_run_cb(void *zhp_data, nvlist_t *nv, void *cb_vcdl)
+for_each_vdev_run_cb(zpool_handle_t *zhp, nvlist_t *nv, void *cb_vcdl)
 {
 	vdev_cmd_data_list_t *vcdl = cb_vcdl;
 	vdev_cmd_data_t *data;
@@ -566,7 +606,6 @@ for_each_vdev_run_cb(void *zhp_data, nvlist_t *nv, void *cb_vcdl)
 	char *vname = NULL;
 	char *vdev_enc_sysfs_path = NULL;
 	int i, match = 0;
-	zpool_handle_t *zhp = zhp_data;
 
 	if (nvlist_lookup_string(nv, ZPOOL_CONFIG_PATH, &path) != 0)
 		return (1);

cmd/zpool/zpool_main.c

@@ -31,8 +31,6 @@
  * Copyright (c) 2017 Open-E, Inc. All Rights Reserved.
  * Copyright (c) 2017, Intel Corporation.
  * Copyright (c) 2019, loli10K <ezomori.nozomu@gmail.com>
- * Copyright (c) 2021, Colm Buckley <colm@tuatha.org>
- * Copyright [2021] Hewlett Packard Enterprise Development LP
  */
 
 #include <assert.h>
@@ -126,9 +124,6 @@ static int zpool_do_version(int, char **);
 
 static int zpool_do_wait(int, char **);
 
-static zpool_compat_status_t zpool_do_load_compat(
-    const char *, boolean_t *);
-
 /*
  * These libumem hooks provide a reasonable set of defaults for the allocator's
  * debugging facilities.
@@ -533,7 +528,7 @@ usage(boolean_t requested)
 		(void) fprintf(fp, "YES   disabled | enabled | active\n");
 
 		(void) fprintf(fp, gettext("\nThe feature@ properties must be "
-		    "appended with a feature name.\nSee zpool-features(7).\n"));
+		    "appended with a feature name.\nSee zpool-features(5).\n"));
 	}
 
 	/*
@@ -787,8 +782,6 @@ add_prop_list(const char *propname, char *propval, nvlist_t **props,
 
 	if (poolprop) {
 		const char *vname = zpool_prop_to_name(ZPOOL_PROP_VERSION);
-		const char *cname =
-		    zpool_prop_to_name(ZPOOL_PROP_COMPATIBILITY);
 
 		if ((prop = zpool_name_to_prop(propname)) == ZPOOL_PROP_INVAL &&
 		    !zpool_prop_feature(propname)) {
@@ -811,22 +804,6 @@ add_prop_list(const char *propname, char *propval, nvlist_t **props,
 			return (2);
 		}
 
-		/*
-		 * if version is specified, only "legacy" compatibility
-		 * may be requested
-		 */
-		if ((prop == ZPOOL_PROP_COMPATIBILITY &&
-		    strcmp(propval, ZPOOL_COMPAT_LEGACY) != 0 &&
-		    nvlist_exists(proplist, vname)) ||
-		    (prop == ZPOOL_PROP_VERSION &&
-		    nvlist_exists(proplist, cname) &&
-		    strcmp(fnvlist_lookup_string(proplist, cname),
-		    ZPOOL_COMPAT_LEGACY) != 0)) {
-			(void) fprintf(stderr, gettext("when 'version' is "
-			    "specified, the 'compatibility' feature may only "
-			    "be set to '" ZPOOL_COMPAT_LEGACY "'\n"));
-			return (2);
-		}
 
 		if (zpool_prop_feature(propname))
 			normnm = propname;
@@ -1069,7 +1046,7 @@ zpool_do_add(int argc, char **argv)
 				free(vname);
 			}
 		}
-		/* And finally the spares */
+		/* And finaly the spares */
 		if (nvlist_lookup_nvlist_array(poolnvroot, ZPOOL_CONFIG_SPARES,
 		    &sparechild, &sparechildren) == 0 && sparechildren > 0) {
 			hadspare = B_TRUE;
@@ -1215,26 +1192,6 @@ zpool_do_remove(int argc, char **argv)
 	return (ret);
 }
 
-/*
- * Return 1 if a vdev is active (being used in a pool)
- * Return 0 if a vdev is inactive (offlined or faulted, or not in active pool)
- *
- * This is useful for checking if a disk in an active pool is offlined or
- * faulted.
- */
-static int
-vdev_is_active(char *vdev_path)
-{
-	int fd;
-	fd = open(vdev_path, O_EXCL);
-	if (fd < 0) {
-		return (1);   /* cant open O_EXCL - disk is active */
-	}
-
-	close(fd);
-	return (0);   /* disk is inactive in the pool */
-}
-
 /*
  * zpool labelclear [-f] <vdev>
  *
@@ -1344,23 +1301,9 @@ zpool_do_labelclear(int argc, char **argv)
 	case POOL_STATE_ACTIVE:
 	case POOL_STATE_SPARE:
 	case POOL_STATE_L2CACHE:
-		/*
-		 * We allow the user to call 'zpool offline -f'
-		 * on an offlined disk in an active pool. We can check if
-		 * the disk is online by calling vdev_is_active().
-		 */
-		if (force && !vdev_is_active(vdev))
-			break;
-
 		(void) fprintf(stderr, gettext(
-		    "%s is a member (%s) of pool \"%s\""),
+		    "%s is a member (%s) of pool \"%s\"\n"),
 		    vdev, zpool_pool_state_to_name(state), name);
-
-		if (force) {
-			(void) fprintf(stderr, gettext(
-			    ". Offline the disk first to clear its label."));
-		}
-		printf("\n");
 		ret = 1;
 		goto errout;
 
@@ -1431,15 +1374,13 @@ zpool_do_create(int argc, char **argv)
 {
 	boolean_t force = B_FALSE;
 	boolean_t dryrun = B_FALSE;
-	boolean_t enable_pool_features = B_TRUE;
+	boolean_t enable_all_pool_feat = B_TRUE;
-
 	int c;
 	nvlist_t *nvroot = NULL;
 	char *poolname;
 	char *tname = NULL;
 	int ret = 1;
 	char *altroot = NULL;
-	char *compat = NULL;
 	char *mountpoint = NULL;
 	nvlist_t *fsprops = NULL;
 	nvlist_t *props = NULL;
@@ -1455,7 +1396,7 @@ zpool_do_create(int argc, char **argv)
 			dryrun = B_TRUE;
 			break;
 		case 'd':
-			enable_pool_features = B_FALSE;
+			enable_all_pool_feat = B_FALSE;
 			break;
 		case 'R':
 			altroot = optarg;
@@ -1493,14 +1434,11 @@ zpool_do_create(int argc, char **argv)
 				ver = strtoull(propval, &end, 10);
 				if (*end == '\0' &&
 				    ver < SPA_VERSION_FEATURES) {
-					enable_pool_features = B_FALSE;
+					enable_all_pool_feat = B_FALSE;
 				}
 			}
 			if (zpool_name_to_prop(optarg) == ZPOOL_PROP_ALTROOT)
 				altroot = propval;
-			if (zpool_name_to_prop(optarg) ==
-			    ZPOOL_PROP_COMPATIBILITY)
-				compat = propval;
 			break;
 		case 'O':
 			if ((propval = strchr(optarg, '=')) == NULL) {
@@ -1694,27 +1632,10 @@ zpool_do_create(int argc, char **argv)
 		ret = 0;
 	} else {
 		/*
-		 * Load in feature set.
+		 * Hand off to libzfs.
-		 * Note: if compatibility property not given, we'll have
-		 * NULL, which means 'all features'.
 		 */
-		boolean_t requested_features[SPA_FEATURES];
+		spa_feature_t i;
-		if (zpool_do_load_compat(compat, requested_features) !=
+		for (i = 0; i < SPA_FEATURES; i++) {
-		    ZPOOL_COMPATIBILITY_OK)
-			goto errout;
-
-		/*
-		 * props contains list of features to enable.
-		 * For each feature:
-		 *  - remove it if feature@name=disabled
-		 *  - leave it there if feature@name=enabled
-		 *  - add it if:
-		 *    - enable_pool_features (ie: no '-d' or '-o version')
-		 *    - it's supported by the kernel module
-		 *    - it's in the requested feature set
-		 *  - warn if it's enabled but not in compat
-		 */
-		for (spa_feature_t i = 0; i < SPA_FEATURES; i++) {
 			char propname[MAXPATHLEN];
 			char *propval;
 			zfeature_info_t *feat = &spa_feature_table[i];
@@ -1722,22 +1643,17 @@ zpool_do_create(int argc, char **argv)
 			(void) snprintf(propname, sizeof (propname),
 			    "feature@%s", feat->fi_uname);
 
+			/*
+			 * Only features contained in props will be enabled:
+			 * remove from the nvlist every ZFS_FEATURE_DISABLED
+			 * value and add every missing ZFS_FEATURE_ENABLED if
+			 * enable_all_pool_feat is set.
+			 */
 			if (!nvlist_lookup_string(props, propname, &propval)) {
 				if (strcmp(propval, ZFS_FEATURE_DISABLED) == 0)
 					(void) nvlist_remove_all(props,
 					    propname);
-				if (strcmp(propval,
+			} else if (enable_all_pool_feat) {
-				    ZFS_FEATURE_ENABLED) == 0 &&
-				    !requested_features[i])
-					(void) fprintf(stderr, gettext(
-					    "Warning: feature \"%s\" enabled "
-					    "but is not in specified "
-					    "'compatibility' feature set.\n"),
-					    feat->fi_uname);
-			} else if (
-			    enable_pool_features &&
-			    feat->fi_zfs_mod_supported &&
-			    requested_features[i]) {
 				ret = add_prop_list(propname,
 				    ZFS_FEATURE_ENABLED, &props, B_TRUE);
 				if (ret != 0)
@@ -2093,7 +2009,7 @@ zpool_print_cmd(vdev_cmd_data_list_t *vcdl, const char *pool, char *path)
 			 * Mark empty values with dashes to make output
 			 * awk-able.
 			 */
-			if (val == NULL || is_blank_str(val))
+			if (is_blank_str(val))
 				val = "-";
 
 			printf("%*s", vcdl->uniq_cols_width[j], val);
@@ -2414,10 +2330,6 @@ print_status_config(zpool_handle_t *zhp, status_cbdata_t *cb, const char *name,
 			(void) printf(gettext("all children offline"));
 			break;
 
-		case VDEV_AUX_BAD_LABEL:
-			(void) printf(gettext("invalid label"));
-			break;
-
 		default:
 			(void) printf(gettext("corrupted data"));
 			break;
@@ -2463,7 +2375,7 @@ print_status_config(zpool_handle_t *zhp, status_cbdata_t *cb, const char *name,
 		}
 	}
 
-	/* Display vdev initialization and trim status for leaves. */
+	/* Display vdev initialization and trim status for leaves */
 	if (children == 0) {
 		print_status_initialize(vs, cb->cb_print_vdev_init);
 		print_status_trim(vs, cb->cb_print_vdev_trim);
@@ -2560,10 +2472,6 @@ print_import_config(status_cbdata_t *cb, const char *name, nvlist_t *nv,
 			(void) printf(gettext("all children offline"));
 			break;
 
-		case VDEV_AUX_BAD_LABEL:
-			(void) printf(gettext("invalid label"));
-			break;
-
 		default:
 			(void) printf(gettext("corrupted data"));
 			break;
@@ -2772,24 +2680,8 @@ show_import(nvlist_t *config, boolean_t report_error)
 
 	case ZPOOL_STATUS_FEAT_DISABLED:
 		printf_color(ANSI_BOLD, gettext("status: "));
-		printf_color(ANSI_YELLOW, gettext("Some supported "
+		printf_color(ANSI_YELLOW, gettext("Some supported features are "
-		    "features are not enabled on the pool.\n\t"
+		    "not enabled on the pool.\n"));
-		    "(Note that they may be intentionally disabled "
-		    "if the\n\t'compatibility' property is set.)\n"));
-		break;
-
-	case ZPOOL_STATUS_COMPATIBILITY_ERR:
-		printf_color(ANSI_BOLD, gettext("status: "));
-		printf_color(ANSI_YELLOW, gettext("Error reading or parsing "
-		    "the file(s) indicated by the 'compatibility'\n"
-		    "property.\n"));
-		break;
-
-	case ZPOOL_STATUS_INCOMPATIBLE_FEAT:
-		printf_color(ANSI_BOLD, gettext("status: "));
-		printf_color(ANSI_YELLOW, gettext("One or more features "
-		    "are enabled on the pool despite not being\n"
-		    "requested by the 'compatibility' property.\n"));
 		break;
 
 	case ZPOOL_STATUS_UNSUP_FEAT_READ:
@@ -2881,12 +2773,6 @@ show_import(nvlist_t *config, boolean_t report_error)
 			    "imported using its name or numeric identifier, "
 			    "though\n\tsome features will not be available "
 			    "without an explicit 'zpool upgrade'.\n"));
-		} else if (reason == ZPOOL_STATUS_COMPATIBILITY_ERR) {
-			(void) printf(gettext(" action: The pool can be "
-			    "imported using its name or numeric\n\tidentifier, "
-			    "though the file(s) indicated by its "
-			    "'compatibility'\n\tproperty cannot be parsed at "
-			    "this time.\n"));
 		} else if (reason == ZPOOL_STATUS_HOSTID_MISMATCH) {
 			(void) printf(gettext(" action: The pool can be "
 			    "imported using its name or numeric "
@@ -3798,10 +3684,9 @@ zpool_do_import(int argc, char **argv)
 		return (1);
 	}
 
-	err = import_pools(pools, props, mntopts, flags,
+	err = import_pools(pools, props, mntopts, flags, argv[0],
-	    argc >= 1 ? argv[0] : NULL,
+	    argc == 1 ? NULL : argv[1], do_destroyed, pool_specified,
-	    argc >= 2 ? argv[1] : NULL,
+	    do_all, &idata);
-	    do_destroyed, pool_specified, do_all, &idata);
 
 	/*
 	 * If we're using the cachefile and we failed to import, then
@@ -3821,10 +3706,9 @@ zpool_do_import(int argc, char **argv)
 		nvlist_free(pools);
 		pools = zpool_search_import(g_zfs, &idata, &libzfs_config_ops);
 
-		err = import_pools(pools, props, mntopts, flags,
+		err = import_pools(pools, props, mntopts, flags, argv[0],
-		    argc >= 1 ? argv[0] : NULL,
+		    argc == 1 ? NULL : argv[1], do_destroyed, pool_specified,
-		    argc >= 2 ? argv[1] : NULL,
+		    do_all, &idata);
-		    do_destroyed, pool_specified, do_all, &idata);
 	}
 
 error:
@@ -4825,7 +4709,7 @@ children:
 			continue;
 
 		vname = zpool_vdev_name(g_zfs, zhp, newchild[c],
-		    cb->cb_name_flags | VDEV_NAME_TYPE_ID);
+		    cb->cb_name_flags);
 		ret += print_vdev_stats(zhp, vname, oldnv ? oldchild[c] : NULL,
 		    newchild[c], cb, depth + 2);
 		free(vname);
@@ -4868,7 +4752,7 @@ children:
 			}
 
 			vname = zpool_vdev_name(g_zfs, zhp, newchild[c],
-			    cb->cb_name_flags | VDEV_NAME_TYPE_ID);
+			    cb->cb_name_flags);
 			ret += print_vdev_stats(zhp, vname, oldnv ?
 			    oldchild[c] : NULL, newchild[c], cb, depth + 2);
 			free(vname);
@@ -5165,12 +5049,11 @@ get_stat_flags(zpool_list_t *list)
  * Return 1 if cb_data->cb_vdev_names[0] is this vdev's name, 0 otherwise.
  */
 static int
-is_vdev_cb(void *zhp_data, nvlist_t *nv, void *cb_data)
+is_vdev_cb(zpool_handle_t *zhp, nvlist_t *nv, void *cb_data)
 {
 	iostat_cbdata_t *cb = cb_data;
 	char *name = NULL;
 	int ret = 0;
-	zpool_handle_t *zhp = zhp_data;
 
 	name = zpool_vdev_name(g_zfs, zhp, nv, cb->cb_name_flags);
 
@@ -6182,7 +6065,7 @@ print_list_stats(zpool_handle_t *zhp, const char *name, nvlist_t *nv,
 			continue;
 
 		vname = zpool_vdev_name(g_zfs, zhp, child[c],
-		    cb->cb_name_flags | VDEV_NAME_TYPE_ID);
+		    cb->cb_name_flags);
 		print_list_stats(zhp, vname, child[c], cb, depth + 2, B_FALSE);
 		free(vname);
 	}
@@ -6216,7 +6099,7 @@ print_list_stats(zpool_handle_t *zhp, const char *name, nvlist_t *nv,
 				printed = B_TRUE;
 			}
 			vname = zpool_vdev_name(g_zfs, zhp, child[c],
-			    cb->cb_name_flags | VDEV_NAME_TYPE_ID);
+			    cb->cb_name_flags);
 			print_list_stats(zhp, vname, child[c], cb, depth + 2,
 			    B_FALSE);
 			free(vname);
@@ -8121,9 +8004,7 @@ status_callback(zpool_handle_t *zhp, void *data)
 	if (cbp->cb_explain &&
 	    (reason == ZPOOL_STATUS_OK ||
 	    reason == ZPOOL_STATUS_VERSION_OLDER ||
-	    reason == ZPOOL_STATUS_FEAT_DISABLED ||
+	    reason == ZPOOL_STATUS_FEAT_DISABLED)) {
-	    reason == ZPOOL_STATUS_COMPATIBILITY_ERR ||
-	    reason == ZPOOL_STATUS_INCOMPATIBLE_FEAT)) {
 		if (!cbp->cb_allpools) {
 			(void) printf(gettext("pool '%s' is healthy\n"),
 			    zpool_get_name(zhp));
@@ -8298,40 +8179,14 @@ status_callback(zpool_handle_t *zhp, void *data)
 
 	case ZPOOL_STATUS_FEAT_DISABLED:
 		printf_color(ANSI_BOLD, gettext("status: "));
-		printf_color(ANSI_YELLOW, gettext("Some supported and "
+		printf_color(ANSI_YELLOW, gettext("Some supported features are "
-		    "requested features are not enabled on the pool.\n\t"
+		    "not enabled on the pool. The pool can\n\tstill be used, "
-		    "The pool can still be used, but some features are "
+		    "but some features are unavailable.\n"));
-		    "unavailable.\n"));
 		printf_color(ANSI_BOLD, gettext("action: "));
 		printf_color(ANSI_YELLOW, gettext("Enable all features using "
 		    "'zpool upgrade'. Once this is done,\n\tthe pool may no "
 		    "longer be accessible by software that does not support\n\t"
-		    "the features. See zpool-features(7) for details.\n"));
+		    "the features. See zpool-features(5) for details.\n"));
-		break;
-
-	case ZPOOL_STATUS_COMPATIBILITY_ERR:
-		printf_color(ANSI_BOLD, gettext("status: "));
-		printf_color(ANSI_YELLOW, gettext("This pool has a "
-		    "compatibility list specified, but it could not be\n\t"
-		    "read/parsed at this time. The pool can still be used, "
-		    "but this\n\tshould be investigated.\n"));
-		printf_color(ANSI_BOLD, gettext("action: "));
-		printf_color(ANSI_YELLOW, gettext("Check the value of the "
-		    "'compatibility' property against the\n\t"
-		    "appropriate file in " ZPOOL_SYSCONF_COMPAT_D " or "
-		    ZPOOL_DATA_COMPAT_D ".\n"));
-		break;
-
-	case ZPOOL_STATUS_INCOMPATIBLE_FEAT:
-		printf_color(ANSI_BOLD, gettext("status: "));
-		printf_color(ANSI_YELLOW, gettext("One or more features "
-		    "are enabled on the pool despite not being\n\t"
-		    "requested by the 'compatibility' property.\n"));
-		printf_color(ANSI_BOLD, gettext("action: "));
-		printf_color(ANSI_YELLOW, gettext("Consider setting "
-		    "'compatibility' to an appropriate value, or\n\t"
-		    "adding needed features to the relevant file in\n\t"
-		    ZPOOL_SYSCONF_COMPAT_D " or " ZPOOL_DATA_COMPAT_D ".\n"));
 		break;
 
 	case ZPOOL_STATUS_UNSUP_FEAT_READ:
@@ -8793,11 +8648,6 @@ upgrade_version(zpool_handle_t *zhp, uint64_t version)
 	verify(nvlist_lookup_uint64(config, ZPOOL_CONFIG_VERSION,
 	    &oldversion) == 0);
 
-	char compat[ZFS_MAXPROPLEN];
-	if (zpool_get_prop(zhp, ZPOOL_PROP_COMPATIBILITY, compat,
-	    ZFS_MAXPROPLEN, NULL, B_FALSE) != 0)
-		compat[0] = '\0';
-
 	assert(SPA_VERSION_IS_SUPPORTED(oldversion));
 	assert(oldversion < version);
 
@@ -8812,13 +8662,6 @@ upgrade_version(zpool_handle_t *zhp, uint64_t version)
 		return (1);
 	}
 
-	if (strcmp(compat, ZPOOL_COMPAT_LEGACY) == 0) {
-		(void) fprintf(stderr, gettext("Upgrade not performed because "
-		    "'compatibility' property set to '"
-		    ZPOOL_COMPAT_LEGACY "'.\n"));
-		return (1);
-	}
-
 	ret = zpool_upgrade(zhp, version);
 	if (ret != 0)
 		return (ret);
@@ -8844,25 +8687,11 @@ upgrade_enable_all(zpool_handle_t *zhp, int *countp)
 	boolean_t firstff = B_TRUE;
 	nvlist_t *enabled = zpool_get_features(zhp);
 
-	char compat[ZFS_MAXPROPLEN];
-	if (zpool_get_prop(zhp, ZPOOL_PROP_COMPATIBILITY, compat,
-	    ZFS_MAXPROPLEN, NULL, B_FALSE) != 0)
-		compat[0] = '\0';
-
-	boolean_t requested_features[SPA_FEATURES];
-	if (zpool_do_load_compat(compat, requested_features) !=
-	    ZPOOL_COMPATIBILITY_OK)
-		return (-1);
-
 	count = 0;
 	for (i = 0; i < SPA_FEATURES; i++) {
 		const char *fname = spa_feature_table[i].fi_uname;
 		const char *fguid = spa_feature_table[i].fi_guid;
-
+		if (!nvlist_exists(enabled, fguid)) {
-		if (!spa_feature_table[i].fi_zfs_mod_supported)
-			continue;
-
-		if (!nvlist_exists(enabled, fguid) && requested_features[i]) {
 			char *propname;
 			verify(-1 != asprintf(&propname, "feature@%s", fname));
 			ret = zpool_set_prop(zhp, propname,
@@ -8895,7 +8724,7 @@ upgrade_cb(zpool_handle_t *zhp, void *arg)
 	upgrade_cbdata_t *cbp = arg;
 	nvlist_t *config;
 	uint64_t version;
-	boolean_t modified_pool = B_FALSE;
+	boolean_t printnl = B_FALSE;
 	int ret;
 
 	config = zpool_get_config(zhp, NULL);
@@ -8909,7 +8738,7 @@ upgrade_cb(zpool_handle_t *zhp, void *arg)
 		ret = upgrade_version(zhp, cbp->cb_version);
 		if (ret != 0)
 			return (ret);
-		modified_pool = B_TRUE;
+		printnl = B_TRUE;
 
 		/*
 		 * If they did "zpool upgrade -a", then we could
@@ -8929,13 +8758,12 @@ upgrade_cb(zpool_handle_t *zhp, void *arg)
 
 		if (count > 0) {
 			cbp->cb_first = B_FALSE;
-			modified_pool = B_TRUE;
+			printnl = B_TRUE;
 		}
 	}
 
-	if (modified_pool) {
+	if (printnl) {
-		(void) printf("\n");
+		(void) printf(gettext("\n"));
-		(void) after_zpool_upgrade(zhp);
 	}
 
 	return (0);
@@ -8961,10 +8789,7 @@ upgrade_list_older_cb(zpool_handle_t *zhp, void *arg)
 			    "be upgraded to use feature flags.  After "
 			    "being upgraded, these pools\nwill no "
 			    "longer be accessible by software that does not "
-			    "support feature\nflags.\n\n"
+			    "support feature\nflags.\n\n"));
-			    "Note that setting a pool's 'compatibility' "
-			    "feature to '" ZPOOL_COMPAT_LEGACY "' will\n"
-			    "inhibit upgrades.\n\n"));
 			(void) printf(gettext("VER  POOL\n"));
 			(void) printf(gettext("---  ------------\n"));
 			cbp->cb_first = B_FALSE;
@@ -8996,10 +8821,6 @@ upgrade_list_disabled_cb(zpool_handle_t *zhp, void *arg)
 		for (i = 0; i < SPA_FEATURES; i++) {
 			const char *fguid = spa_feature_table[i].fi_guid;
 			const char *fname = spa_feature_table[i].fi_uname;
-
-			if (!spa_feature_table[i].fi_zfs_mod_supported)
-				continue;
-
 			if (!nvlist_exists(enabled, fguid)) {
 				if (cbp->cb_first) {
 					(void) printf(gettext("\nSome "
@@ -9009,12 +8830,8 @@ upgrade_list_disabled_cb(zpool_handle_t *zhp, void *arg)
 					    "pool may become incompatible with "
 					    "software\nthat does not support "
 					    "the feature. See "
-					    "zpool-features(7) for "
+					    "zpool-features(5) for "
-					    "details.\n\n"
+					    "details.\n\n"));
-					    "Note that the pool "
-					    "'compatibility' feature can be "
-					    "used to inhibit\nfeature "
-					    "upgrades.\n\n"));
 					(void) printf(gettext("POOL  "
 					    "FEATURE\n"));
 					(void) printf(gettext("------"
@@ -9048,7 +8865,7 @@ upgrade_list_disabled_cb(zpool_handle_t *zhp, void *arg)
 static int
 upgrade_one(zpool_handle_t *zhp, void *data)
 {
-	boolean_t modified_pool = B_FALSE;
+	boolean_t printnl = B_FALSE;
 	upgrade_cbdata_t *cbp = data;
 	uint64_t cur_version;
 	int ret;
@@ -9076,7 +8893,7 @@ upgrade_one(zpool_handle_t *zhp, void *data)
 	}
 
 	if (cur_version != cbp->cb_version) {
-		modified_pool = B_TRUE;
+		printnl = B_TRUE;
 		ret = upgrade_version(zhp, cbp->cb_version);
 		if (ret != 0)
 			return (ret);
@@ -9089,17 +8906,16 @@ upgrade_one(zpool_handle_t *zhp, void *data)
 			return (ret);
 
 		if (count != 0) {
-			modified_pool = B_TRUE;
+			printnl = B_TRUE;
 		} else if (cur_version == SPA_VERSION) {
 			(void) printf(gettext("Pool '%s' already has all "
-			    "supported and requested features enabled.\n"),
+			    "supported features enabled.\n"),
 			    zpool_get_name(zhp));
 		}
 	}
 
-	if (modified_pool) {
+	if (printnl) {
-		(void) printf("\n");
+		(void) printf(gettext("\n"));
-		(void) after_zpool_upgrade(zhp);
 	}
 
 	return (0);
@@ -9194,8 +9010,6 @@ zpool_do_upgrade(int argc, char **argv)
 		    "---------------\n");
 		for (i = 0; i < SPA_FEATURES; i++) {
 			zfeature_info_t *fi = &spa_feature_table[i];
-			if (!fi->fi_zfs_mod_supported)
-				continue;
 			const char *ro =
 			    (fi->fi_flags & ZFEATURE_FLAG_READONLY_COMPAT) ?
 			    " (read-only compatible)" : "";
@@ -9256,8 +9070,8 @@ zpool_do_upgrade(int argc, char **argv)
 				(void) printf(gettext("All pools are already "
 				    "formatted using feature flags.\n\n"));
 				(void) printf(gettext("Every feature flags "
-				    "pool already has all supported and "
+				    "pool already has all supported features "
-				    "requested features enabled.\n"));
+				    "enabled.\n"));
 			} else {
 				(void) printf(gettext("All pools are already "
 				    "formatted with version %llu or higher.\n"),
@@ -9283,7 +9097,7 @@ zpool_do_upgrade(int argc, char **argv)
 
 		if (cb.cb_first) {
 			(void) printf(gettext("Every feature flags pool has "
-			    "all supported and requested features enabled.\n"));
+			    "all supported features enabled.\n"));
 		} else {
 			(void) printf(gettext("\n"));
 		}
@@ -9312,7 +9126,7 @@ print_history_records(nvlist_t *nvhis, hist_cbdata_t *cb)
 	    &records, &numrecords) == 0);
 	for (i = 0; i < numrecords; i++) {
 		nvlist_t *rec = records[i];
-		char tbuf[64] = "";
+		char tbuf[30] = "";
 
 		if (nvlist_exists(rec, ZPOOL_HIST_TIME)) {
 			time_t tsec;
@@ -9324,14 +9138,6 @@ print_history_records(nvlist_t *nvhis, hist_cbdata_t *cb)
 			(void) strftime(tbuf, sizeof (tbuf), "%F.%T", &t);
 		}
 
-		if (nvlist_exists(rec, ZPOOL_HIST_ELAPSED_NS)) {
-			uint64_t elapsed_ns = fnvlist_lookup_int64(records[i],
-			    ZPOOL_HIST_ELAPSED_NS);
-			(void) snprintf(tbuf + strlen(tbuf),
-			    sizeof (tbuf) - strlen(tbuf),
-			    " (%lldms)", (long long)elapsed_ns / 1000 / 1000);
-		}
-
 		if (nvlist_exists(rec, ZPOOL_HIST_CMD)) {
 			(void) printf("%s %s", tbuf,
 			    fnvlist_lookup_string(rec, ZPOOL_HIST_CMD));
@@ -10071,63 +9877,6 @@ set_callback(zpool_handle_t *zhp, void *data)
 	int error;
 	set_cbdata_t *cb = (set_cbdata_t *)data;
 
-	/* Check if we have out-of-bounds features */
-	if (strcmp(cb->cb_propname, ZPOOL_CONFIG_COMPATIBILITY) == 0) {
-		boolean_t features[SPA_FEATURES];
-		if (zpool_do_load_compat(cb->cb_value, features) !=
-		    ZPOOL_COMPATIBILITY_OK)
-			return (-1);
-
-		nvlist_t *enabled = zpool_get_features(zhp);
-		spa_feature_t i;
-		for (i = 0; i < SPA_FEATURES; i++) {
-			const char *fguid = spa_feature_table[i].fi_guid;
-			if (nvlist_exists(enabled, fguid) && !features[i])
-				break;
-		}
-		if (i < SPA_FEATURES)
-			(void) fprintf(stderr, gettext("Warning: one or "
-			    "more features already enabled on pool '%s'\n"
-			    "are not present in this compatibility set.\n"),
-			    zpool_get_name(zhp));
-	}
-
-	/* if we're setting a feature, check it's in compatibility set */
-	if (zpool_prop_feature(cb->cb_propname) &&
-	    strcmp(cb->cb_value, ZFS_FEATURE_ENABLED) == 0) {
-		char *fname = strchr(cb->cb_propname, '@') + 1;
-		spa_feature_t f;
-
-		if (zfeature_lookup_name(fname, &f) == 0) {
-			char compat[ZFS_MAXPROPLEN];
-			if (zpool_get_prop(zhp, ZPOOL_PROP_COMPATIBILITY,
-			    compat, ZFS_MAXPROPLEN, NULL, B_FALSE) != 0)
-				compat[0] = '\0';
-
-			boolean_t features[SPA_FEATURES];
-			if (zpool_do_load_compat(compat, features) !=
-			    ZPOOL_COMPATIBILITY_OK) {
-				(void) fprintf(stderr, gettext("Error: "
-				    "cannot enable feature '%s' on pool '%s'\n"
-				    "because the pool's 'compatibility' "
-				    "property cannot be parsed.\n"),
-				    fname, zpool_get_name(zhp));
-				return (-1);
-			}
-
-			if (!features[f]) {
-				(void) fprintf(stderr, gettext("Error: "
-				    "cannot enable feature '%s' on pool '%s'\n"
-				    "as it is not specified in this pool's "
-				    "current compatibility set.\n"
-				    "Consider setting 'compatibility' to a "
-				    "less restrictive set, or to 'off'.\n"),
-				    fname, zpool_get_name(zhp));
-				return (-1);
-			}
-		}
-	}
-
 	error = zpool_set_prop(zhp, cb->cb_propname, cb->cb_value);
 
 	if (!error)
@@ -10258,8 +10007,7 @@ vdev_any_spare_replacing(nvlist_t *nv)
 	(void) nvlist_lookup_string(nv, ZPOOL_CONFIG_TYPE, &vdev_type);
 
 	if (strcmp(vdev_type, VDEV_TYPE_REPLACING) == 0 ||
-	    strcmp(vdev_type, VDEV_TYPE_SPARE) == 0 ||
+	    strcmp(vdev_type, VDEV_TYPE_SPARE) == 0) {
-	    strcmp(vdev_type, VDEV_TYPE_DRAID_SPARE) == 0) {
 		return (B_TRUE);
 	}
 
@@ -10644,36 +10392,6 @@ zpool_do_version(int argc, char **argv)
 	return (0);
 }
 
-/*
- * Do zpool_load_compat() and print error message on failure
- */
-static zpool_compat_status_t
-zpool_do_load_compat(const char *compat, boolean_t *list)
-{
-	char report[1024];
-
-	zpool_compat_status_t ret;
-
-	ret = zpool_load_compat(compat, list, report, 1024);
-	switch (ret) {
-
-	case ZPOOL_COMPATIBILITY_OK:
-		break;
-
-	case ZPOOL_COMPATIBILITY_NOFILES:
-	case ZPOOL_COMPATIBILITY_BADFILE:
-	case ZPOOL_COMPATIBILITY_BADTOKEN:
-		(void) fprintf(stderr, "Error: %s\n", report);
-		break;
-
-	case ZPOOL_COMPATIBILITY_WARNTOKEN:
-		(void) fprintf(stderr, "Warning: %s\n", report);
-		ret = ZPOOL_COMPATIBILITY_OK;
-		break;
-	}
-	return (ret);
-}
-
 int
 main(int argc, char **argv)
 {

cmd/zpool/zpool_util.h

@@ -27,7 +27,6 @@
 
 #include <libnvpair.h>
 #include <libzfs.h>
-#include <libzutil.h>
 
 #ifdef	__cplusplus
 extern "C" {
@@ -68,6 +67,7 @@ int for_each_pool(int, char **, boolean_t unavail, zprop_list_t **,
     boolean_t, zpool_iter_f, void *);
 
 /* Vdev list functions */
+typedef int (*pool_vdev_iter_f)(zpool_handle_t *, nvlist_t *, void *);
 int for_each_vdev(zpool_handle_t *zhp, pool_vdev_iter_f func, void *data);
 
 typedef struct zpool_list zpool_list_t;
@@ -129,7 +129,6 @@ int check_device(const char *path, boolean_t force,
 boolean_t check_sector_size_database(char *path, int *sector_size);
 void vdev_error(const char *fmt, ...);
 int check_file(const char *file, boolean_t force, boolean_t isspare);
-void after_zpool_upgrade(zpool_handle_t *zhp);
 
 #ifdef	__cplusplus
 }

cmd/zpool/zpool_vdev.c

@@ -86,6 +86,9 @@
 boolean_t error_seen;
 boolean_t is_force;
 
+
+
+
 /*PRINTFLIKE1*/
 void
 vdev_error(const char *fmt, ...)
@@ -219,9 +222,6 @@ is_spare(nvlist_t *config, const char *path)
 	uint_t i, nspares;
 	boolean_t inuse;
 
-	if (zpool_is_draid_spare(path))
-		return (B_TRUE);
-
 	if ((fd = open(path, O_RDONLY|O_DIRECT)) < 0)
 		return (B_FALSE);
 
@@ -267,10 +267,9 @@ is_spare(nvlist_t *config, const char *path)
  *	/dev/xxx	Complete disk path
  *	/xxx		Full path to file
  *	xxx		Shorthand for <zfs_vdev_paths>/xxx
- *	draid*		Virtual dRAID spare
  */
 static nvlist_t *
-make_leaf_vdev(nvlist_t *props, const char *arg, boolean_t is_primary)
+make_leaf_vdev(nvlist_t *props, const char *arg, uint64_t is_log)
 {
 	char path[MAXPATHLEN];
 	struct stat64 statbuf;
@@ -310,17 +309,6 @@ make_leaf_vdev(nvlist_t *props, const char *arg, boolean_t is_primary)
 
 		/* After whole disk check restore original passed path */
 		strlcpy(path, arg, sizeof (path));
-	} else if (zpool_is_draid_spare(arg)) {
-		if (!is_primary) {
-			(void) fprintf(stderr,
-			    gettext("cannot open '%s': dRAID spares can only "
-			    "be used to replace primary vdevs\n"), arg);
-			return (NULL);
-		}
-
-		wholedisk = B_TRUE;
-		strlcpy(path, arg, sizeof (path));
-		type = VDEV_TYPE_DRAID_SPARE;
 	} else {
 		err = is_shorthand_path(arg, path, sizeof (path),
 		    &statbuf, &wholedisk);
@@ -349,19 +337,17 @@ make_leaf_vdev(nvlist_t *props, const char *arg, boolean_t is_primary)
 		}
 	}
 
-	if (type == NULL) {
+	/*
-		/*
+	 * Determine whether this is a device or a file.
-		 * Determine whether this is a device or a file.
+	 */
-		 */
+	if (wholedisk || S_ISBLK(statbuf.st_mode)) {
-		if (wholedisk || S_ISBLK(statbuf.st_mode)) {
+		type = VDEV_TYPE_DISK;
-			type = VDEV_TYPE_DISK;
+	} else if (S_ISREG(statbuf.st_mode)) {
-		} else if (S_ISREG(statbuf.st_mode)) {
+		type = VDEV_TYPE_FILE;
-			type = VDEV_TYPE_FILE;
+	} else {
-		} else {
+		(void) fprintf(stderr, gettext("cannot use '%s': must be a "
-			fprintf(stderr, gettext("cannot use '%s': must "
+		    "block device or regular file\n"), path);
-			    "be a block device or regular file\n"), path);
+		return (NULL);
-			return (NULL);
-		}
 	}
 
 	/*
@@ -372,7 +358,10 @@ make_leaf_vdev(nvlist_t *props, const char *arg, boolean_t is_primary)
 	verify(nvlist_alloc(&vdev, NV_UNIQUE_NAME, 0) == 0);
 	verify(nvlist_add_string(vdev, ZPOOL_CONFIG_PATH, path) == 0);
 	verify(nvlist_add_string(vdev, ZPOOL_CONFIG_TYPE, type) == 0);
-
+	verify(nvlist_add_uint64(vdev, ZPOOL_CONFIG_IS_LOG, is_log) == 0);
+	if (is_log)
+		verify(nvlist_add_string(vdev, ZPOOL_CONFIG_ALLOCATION_BIAS,
+		    VDEV_ALLOC_BIAS_LOG) == 0);
 	if (strcmp(type, VDEV_TYPE_DISK) == 0)
 		verify(nvlist_add_uint64(vdev, ZPOOL_CONFIG_WHOLE_DISK,
 		    (uint64_t)wholedisk) == 0);
@@ -443,16 +432,11 @@ typedef struct replication_level {
 
 #define	ZPOOL_FUZZ	(16 * 1024 * 1024)
 
-/*
- * N.B. For the purposes of comparing replication levels dRAID can be
- * considered functionally equivalent to raidz.
- */
 static boolean_t
 is_raidz_mirror(replication_level_t *a, replication_level_t *b,
     replication_level_t **raidz, replication_level_t **mirror)
 {
-	if ((strcmp(a->zprl_type, "raidz") == 0 ||
+	if (strcmp(a->zprl_type, "raidz") == 0 &&
-	    strcmp(a->zprl_type, "draid") == 0) &&
 	    strcmp(b->zprl_type, "mirror") == 0) {
 		*raidz = a;
 		*mirror = b;
@@ -461,22 +445,6 @@ is_raidz_mirror(replication_level_t *a, replication_level_t *b,
 	return (B_FALSE);
 }
 
-/*
- * Comparison for determining if dRAID and raidz where passed in either order.
- */
-static boolean_t
-is_raidz_draid(replication_level_t *a, replication_level_t *b)
-{
-	if ((strcmp(a->zprl_type, "raidz") == 0 ||
-	    strcmp(a->zprl_type, "draid") == 0) &&
-	    (strcmp(b->zprl_type, "raidz") == 0 ||
-	    strcmp(b->zprl_type, "draid") == 0)) {
-		return (B_TRUE);
-	}
-
-	return (B_FALSE);
-}
-
 /*
  * Given a list of toplevel vdevs, return the current replication level.  If
  * the config is inconsistent, then NULL is returned.  If 'fatal' is set, then
@@ -543,8 +511,7 @@ get_replication(nvlist_t *nvroot, boolean_t fatal)
 			rep.zprl_type = type;
 			rep.zprl_children = 0;
 
-			if (strcmp(type, VDEV_TYPE_RAIDZ) == 0 ||
+			if (strcmp(type, VDEV_TYPE_RAIDZ) == 0) {
-			    strcmp(type, VDEV_TYPE_DRAID) == 0) {
 				verify(nvlist_lookup_uint64(nv,
 				    ZPOOL_CONFIG_NPARITY,
 				    &rep.zprl_parity) == 0);
@@ -710,29 +677,6 @@ get_replication(nvlist_t *nvroot, boolean_t fatal)
 					else
 						return (NULL);
 				}
-			} else if (is_raidz_draid(&lastrep, &rep)) {
-				/*
-				 * Accepted raidz and draid when they can
-				 * handle the same number of disk failures.
-				 */
-				if (lastrep.zprl_parity != rep.zprl_parity) {
-					if (ret != NULL)
-						free(ret);
-					ret = NULL;
-					if (fatal)
-						vdev_error(gettext(
-						    "mismatched replication "
-						    "level: %s and %s vdevs "
-						    "with different "
-						    "redundancy, %llu vs. "
-						    "%llu are present\n"),
-						    lastrep.zprl_type,
-						    rep.zprl_type,
-						    lastrep.zprl_parity,
-						    rep.zprl_parity);
-					else
-						return (NULL);
-				}
 			} else if (strcmp(lastrep.zprl_type, rep.zprl_type) !=
 			    0) {
 				if (ret != NULL)
@@ -1159,87 +1103,31 @@ is_device_in_use(nvlist_t *config, nvlist_t *nv, boolean_t force,
 	return (anyinuse);
 }
 
-/*
- * Returns the parity level extracted from a raidz or draid type.
- * If the parity cannot be determined zero is returned.
- */
-static int
-get_parity(const char *type)
-{
-	long parity = 0;
-	const char *p;
-
-	if (strncmp(type, VDEV_TYPE_RAIDZ, strlen(VDEV_TYPE_RAIDZ)) == 0) {
-		p = type + strlen(VDEV_TYPE_RAIDZ);
-
-		if (*p == '\0') {
-			/* when unspecified default to single parity */
-			return (1);
-		} else if (*p == '0') {
-			/* no zero prefixes allowed */
-			return (0);
-		} else {
-			/* 0-3, no suffixes allowed */
-			char *end;
-			errno = 0;
-			parity = strtol(p, &end, 10);
-			if (errno != 0 || *end != '\0' ||
-			    parity < 1 || parity > VDEV_RAIDZ_MAXPARITY) {
-				return (0);
-			}
-		}
-	} else if (strncmp(type, VDEV_TYPE_DRAID,
-	    strlen(VDEV_TYPE_DRAID)) == 0) {
-		p = type + strlen(VDEV_TYPE_DRAID);
-
-		if (*p == '\0' || *p == ':') {
-			/* when unspecified default to single parity */
-			return (1);
-		} else if (*p == '0') {
-			/* no zero prefixes allowed */
-			return (0);
-		} else {
-			/* 0-3, allowed suffixes: '\0' or ':' */
-			char *end;
-			errno = 0;
-			parity = strtol(p, &end, 10);
-			if (errno != 0 ||
-			    parity < 1 || parity > VDEV_DRAID_MAXPARITY ||
-			    (*end != '\0' && *end != ':')) {
-				return (0);
-			}
-		}
-	}
-
-	return ((int)parity);
-}
-
-/*
- * Assign the minimum and maximum number of devices allowed for
- * the specified type.  On error NULL is returned, otherwise the
- * type prefix is returned (raidz, mirror, etc).
- */
 static const char *
 is_grouping(const char *type, int *mindev, int *maxdev)
 {
-	int nparity;
+	if (strncmp(type, "raidz", 5) == 0) {
+		const char *p = type + 5;
+		char *end;
+		long nparity;
+
+		if (*p == '\0') {
+			nparity = 1;
+		} else if (*p == '0') {
+			return (NULL); /* no zero prefixes allowed */
+		} else {
+			errno = 0;
+			nparity = strtol(p, &end, 10);
+			if (errno != 0 || nparity < 1 || nparity >= 255 ||
+			    *end != '\0')
+				return (NULL);
+		}
 
-	if (strncmp(type, VDEV_TYPE_RAIDZ, strlen(VDEV_TYPE_RAIDZ)) == 0 ||
-	    strncmp(type, VDEV_TYPE_DRAID, strlen(VDEV_TYPE_DRAID)) == 0) {
-		nparity = get_parity(type);
-		if (nparity == 0)
-			return (NULL);
 		if (mindev != NULL)
 			*mindev = nparity + 1;
 		if (maxdev != NULL)
 			*maxdev = 255;
-
+		return (VDEV_TYPE_RAIDZ);
-		if (strncmp(type, VDEV_TYPE_RAIDZ,
-		    strlen(VDEV_TYPE_RAIDZ)) == 0) {
-			return (VDEV_TYPE_RAIDZ);
-		} else {
-			return (VDEV_TYPE_DRAID);
-		}
 	}
 
 	if (maxdev != NULL)
@@ -1279,163 +1167,6 @@ is_grouping(const char *type, int *mindev, int *maxdev)
 	return (NULL);
 }
 
-/*
- * Extract the configuration parameters encoded in the dRAID type and
- * use them to generate a dRAID configuration.  The expected format is:
- *
- * draid[<parity>][:<data><d|D>][:<children><c|C>][:<spares><s|S>]
- *
- * The intent is to be able to generate a good configuration when no
- * additional information is provided.  The only mandatory component
- * of the 'type' is the 'draid' prefix.  If a value is not provided
- * then reasonable defaults are used.  The optional components may
- * appear in any order but the d/s/c suffix is required.
- *
- * Valid inputs:
- * - data:     number of data devices per group (1-255)
- * - parity:   number of parity blocks per group (1-3)
- * - spares:   number of distributed spare (0-100)
- * - children: total number of devices (1-255)
- *
- * Examples:
- * - zpool create tank draid <devices...>
- * - zpool create tank draid2:8d:51c:2s <devices...>
- */
-static int
-draid_config_by_type(nvlist_t *nv, const char *type, uint64_t children)
-{
-	uint64_t nparity = 1;
-	uint64_t nspares = 0;
-	uint64_t ndata = UINT64_MAX;
-	uint64_t ngroups = 1;
-	long value;
-
-	if (strncmp(type, VDEV_TYPE_DRAID, strlen(VDEV_TYPE_DRAID)) != 0)
-		return (EINVAL);
-
-	nparity = (uint64_t)get_parity(type);
-	if (nparity == 0)
-		return (EINVAL);
-
-	char *p = (char *)type;
-	while ((p = strchr(p, ':')) != NULL) {
-		char *end;
-
-		p = p + 1;
-		errno = 0;
-
-		if (!isdigit(p[0])) {
-			(void) fprintf(stderr, gettext("invalid dRAID "
-			    "syntax; expected [:<number><c|d|s>] not '%s'\n"),
-			    type);
-			return (EINVAL);
-		}
-
-		/* Expected non-zero value with c/d/s suffix */
-		value = strtol(p, &end, 10);
-		char suffix = tolower(*end);
-		if (errno != 0 ||
-		    (suffix != 'c' && suffix != 'd' && suffix != 's')) {
-			(void) fprintf(stderr, gettext("invalid dRAID "
-			    "syntax; expected [:<number><c|d|s>] not '%s'\n"),
-			    type);
-			return (EINVAL);
-		}
-
-		if (suffix == 'c') {
-			if ((uint64_t)value != children) {
-				fprintf(stderr,
-				    gettext("invalid number of dRAID children; "
-				    "%llu required but %llu provided\n"),
-				    (u_longlong_t)value,
-				    (u_longlong_t)children);
-				return (EINVAL);
-			}
-		} else if (suffix == 'd') {
-			ndata = (uint64_t)value;
-		} else if (suffix == 's') {
-			nspares = (uint64_t)value;
-		} else {
-			verify(0); /* Unreachable */
-		}
-	}
-
-	/*
-	 * When a specific number of data disks is not provided limit a
-	 * redundancy group to 8 data disks.  This value was selected to
-	 * provide a reasonable tradeoff between capacity and performance.
-	 */
-	if (ndata == UINT64_MAX) {
-		if (children > nspares + nparity) {
-			ndata = MIN(children - nspares - nparity, 8);
-		} else {
-			fprintf(stderr, gettext("request number of "
-			    "distributed spares %llu and parity level %llu\n"
-			    "leaves no disks available for data\n"),
-			    (u_longlong_t)nspares, (u_longlong_t)nparity);
-			return (EINVAL);
-		}
-	}
-
-	/* Verify the maximum allowed group size is never exceeded. */
-	if (ndata == 0 || (ndata + nparity > children - nspares)) {
-		fprintf(stderr, gettext("requested number of dRAID data "
-		    "disks per group %llu is too high,\nat most %llu disks "
-		    "are available for data\n"), (u_longlong_t)ndata,
-		    (u_longlong_t)(children - nspares - nparity));
-		return (EINVAL);
-	}
-
-	if (nparity == 0 || nparity > VDEV_DRAID_MAXPARITY) {
-		fprintf(stderr,
-		    gettext("invalid dRAID parity level %llu; must be "
-		    "between 1 and %d\n"), (u_longlong_t)nparity,
-		    VDEV_DRAID_MAXPARITY);
-		return (EINVAL);
-	}
-
-	/*
-	 * Verify the requested number of spares can be satisfied.
-	 * An arbitrary limit of 100 distributed spares is applied.
-	 */
-	if (nspares > 100 || nspares > (children - (ndata + nparity))) {
-		fprintf(stderr,
-		    gettext("invalid number of dRAID spares %llu; additional "
-		    "disks would be required\n"), (u_longlong_t)nspares);
-		return (EINVAL);
-	}
-
-	/* Verify the requested number children is sufficient. */
-	if (children < (ndata + nparity + nspares)) {
-		fprintf(stderr, gettext("%llu disks were provided, but at "
-		    "least %llu disks are required for this config\n"),
-		    (u_longlong_t)children,
-		    (u_longlong_t)(ndata + nparity + nspares));
-	}
-
-	if (children > VDEV_DRAID_MAX_CHILDREN) {
-		fprintf(stderr, gettext("%llu disks were provided, but "
-		    "dRAID only supports up to %u disks"),
-		    (u_longlong_t)children, VDEV_DRAID_MAX_CHILDREN);
-	}
-
-	/*
-	 * Calculate the minimum number of groups required to fill a slice.
-	 * This is the LCM of the stripe width (ndata + nparity) and the
-	 * number of data drives (children - nspares).
-	 */
-	while (ngroups * (ndata + nparity) % (children - nspares) != 0)
-		ngroups++;
-
-	/* Store the basic dRAID configuration. */
-	fnvlist_add_uint64(nv, ZPOOL_CONFIG_NPARITY, nparity);
-	fnvlist_add_uint64(nv, ZPOOL_CONFIG_DRAID_NDATA, ndata);
-	fnvlist_add_uint64(nv, ZPOOL_CONFIG_DRAID_NSPARES, nspares);
-	fnvlist_add_uint64(nv, ZPOOL_CONFIG_DRAID_NGROUPS, ngroups);
-
-	return (0);
-}
-
 /*
  * Construct a syntactically valid vdev specification,
  * and ensure that all devices and files exist and can be opened.
@@ -1447,8 +1178,8 @@ construct_spec(nvlist_t *props, int argc, char **argv)
 {
 	nvlist_t *nvroot, *nv, **top, **spares, **l2cache;
 	int t, toplevels, mindev, maxdev, nspares, nlogs, nl2cache;
-	const char *type, *fulltype;
+	const char *type;
-	boolean_t is_log, is_special, is_dedup, is_spare;
+	uint64_t is_log, is_special, is_dedup;
 	boolean_t seen_logs;
 
 	top = NULL;
@@ -1458,20 +1189,18 @@ construct_spec(nvlist_t *props, int argc, char **argv)
 	nspares = 0;
 	nlogs = 0;
 	nl2cache = 0;
-	is_log = is_special = is_dedup = is_spare = B_FALSE;
+	is_log = is_special = is_dedup = B_FALSE;
 	seen_logs = B_FALSE;
 	nvroot = NULL;
 
 	while (argc > 0) {
-		fulltype = argv[0];
 		nv = NULL;
 
 		/*
-		 * If it's a mirror, raidz, or draid the subsequent arguments
+		 * If it's a mirror or raidz, the subsequent arguments are
-		 * are its leaves -- until we encounter the next mirror,
+		 * its leaves -- until we encounter the next mirror or raidz.
-		 * raidz or draid.
 		 */
-		if ((type = is_grouping(fulltype, &mindev, &maxdev)) != NULL) {
+		if ((type = is_grouping(argv[0], &mindev, &maxdev)) != NULL) {
 			nvlist_t **child = NULL;
 			int c, children = 0;
 
@@ -1483,7 +1212,6 @@ construct_spec(nvlist_t *props, int argc, char **argv)
 					    "specified only once\n"));
 					goto spec_out;
 				}
-				is_spare = B_TRUE;
 				is_log = is_special = is_dedup = B_FALSE;
 			}
 
@@ -1497,7 +1225,8 @@ construct_spec(nvlist_t *props, int argc, char **argv)
 				}
 				seen_logs = B_TRUE;
 				is_log = B_TRUE;
-				is_special = is_dedup = is_spare = B_FALSE;
+				is_special = B_FALSE;
+				is_dedup = B_FALSE;
 				argc--;
 				argv++;
 				/*
@@ -1509,7 +1238,8 @@ construct_spec(nvlist_t *props, int argc, char **argv)
 
 			if (strcmp(type, VDEV_ALLOC_BIAS_SPECIAL) == 0) {
 				is_special = B_TRUE;
-				is_log = is_dedup = is_spare = B_FALSE;
+				is_log = B_FALSE;
+				is_dedup = B_FALSE;
 				argc--;
 				argv++;
 				continue;
@@ -1517,7 +1247,8 @@ construct_spec(nvlist_t *props, int argc, char **argv)
 
 			if (strcmp(type, VDEV_ALLOC_BIAS_DEDUP) == 0) {
 				is_dedup = B_TRUE;
-				is_log = is_special = is_spare = B_FALSE;
+				is_log = B_FALSE;
+				is_special = B_FALSE;
 				argc--;
 				argv++;
 				continue;
@@ -1531,8 +1262,7 @@ construct_spec(nvlist_t *props, int argc, char **argv)
 					    "specified only once\n"));
 					goto spec_out;
 				}
-				is_log = is_special = B_FALSE;
+				is_log = is_special = is_dedup = B_FALSE;
-				is_dedup = is_spare = B_FALSE;
 			}
 
 			if (is_log || is_special || is_dedup) {
@@ -1550,15 +1280,13 @@ construct_spec(nvlist_t *props, int argc, char **argv)
 			for (c = 1; c < argc; c++) {
 				if (is_grouping(argv[c], NULL, NULL) != NULL)
 					break;
-
 				children++;
 				child = realloc(child,
 				    children * sizeof (nvlist_t *));
 				if (child == NULL)
 					zpool_no_memory();
 				if ((nv = make_leaf_vdev(props, argv[c],
-				    !(is_log || is_special || is_dedup ||
+				    B_FALSE)) == NULL) {
-				    is_spare))) == NULL) {
 					for (c = 0; c < children - 1; c++)
 						nvlist_free(child[c]);
 					free(child);
@@ -1607,11 +1335,10 @@ construct_spec(nvlist_t *props, int argc, char **argv)
 				    type) == 0);
 				verify(nvlist_add_uint64(nv,
 				    ZPOOL_CONFIG_IS_LOG, is_log) == 0);
-				if (is_log) {
+				if (is_log)
 					verify(nvlist_add_string(nv,
 					    ZPOOL_CONFIG_ALLOCATION_BIAS,
 					    VDEV_ALLOC_BIAS_LOG) == 0);
-				}
 				if (is_special) {
 					verify(nvlist_add_string(nv,
 					    ZPOOL_CONFIG_ALLOCATION_BIAS,
@@ -1627,15 +1354,6 @@ construct_spec(nvlist_t *props, int argc, char **argv)
 					    ZPOOL_CONFIG_NPARITY,
 					    mindev - 1) == 0);
 				}
-				if (strcmp(type, VDEV_TYPE_DRAID) == 0) {
-					if (draid_config_by_type(nv,
-					    fulltype, children) != 0) {
-						for (c = 0; c < children; c++)
-							nvlist_free(child[c]);
-						free(child);
-						goto spec_out;
-					}
-				}
 				verify(nvlist_add_nvlist_array(nv,
 				    ZPOOL_CONFIG_CHILDREN, child,
 				    children) == 0);
@@ -1649,19 +1367,12 @@ construct_spec(nvlist_t *props, int argc, char **argv)
 			 * We have a device.  Pass off to make_leaf_vdev() to
 			 * construct the appropriate nvlist describing the vdev.
 			 */
-			if ((nv = make_leaf_vdev(props, argv[0], !(is_log ||
+			if ((nv = make_leaf_vdev(props, argv[0],
-			    is_special || is_dedup || is_spare))) == NULL)
+			    is_log)) == NULL)
 				goto spec_out;
 
-			verify(nvlist_add_uint64(nv,
+			if (is_log)
-			    ZPOOL_CONFIG_IS_LOG, is_log) == 0);
-			if (is_log) {
-				verify(nvlist_add_string(nv,
-				    ZPOOL_CONFIG_ALLOCATION_BIAS,
-				    VDEV_ALLOC_BIAS_LOG) == 0);
 				nlogs++;
-			}
-
 			if (is_special) {
 				verify(nvlist_add_string(nv,
 				    ZPOOL_CONFIG_ALLOCATION_BIAS,

cmd/zpool_influxdb/.gitignore

@@ -1 +0,0 @@
-/zpool_influxdb