Add pool state /proc entry, "SUSPENDED" pools

1. Add a proc entry to display the pool's state:

$ cat /proc/spl/kstat/zfs/tank/state
ONLINE

This is done without using the spa config locks, so it will
never hang.

2. Fix 'zpool status' and 'zpool list -o health' output to print
"SUSPENDED" instead of "ONLINE" for suspended pools.

Reviewed-by: Olaf Faaland <[email protected]>
Reviewed-by: Brian Behlendorf <[email protected]>
Reviewed by: Richard Elling <[email protected]>
Signed-off-by: Tony Hutter <[email protected]>
Closes #7331 
Closes #7563 

6 files changed, 222 insertions, 1 deletions

diff --git a/tests/zfs-tests/include/blkdev.shlib b/tests/zfs-tests/include/blkdev.shlib
index 87ffa8560..5163ea2ae 100644
--- a/tests/zfs-tests/include/blkdev.shlib
+++ b/tests/zfs-tests/include/blkdev.shlib
@@ -421,7 +421,16 @@ function unload_scsi_debug
 #
 function get_debug_device
 {
-	lsscsi | nawk '/scsi_debug/ {print $6; exit}' | cut -d / -f3
+	for i in {1..10} ; do
+		val=$(lsscsi | nawk '/scsi_debug/ {print $6; exit}' | cut -d / -f3)
+
+		# lsscsi can take time to settle
+		if [ "$val" != "-" ] ; then
+			break
+		fi
+		sleep 1
+	done
+	echo "$val"
 }
 
 #
diff --git a/tests/zfs-tests/tests/functional/Makefile.am b/tests/zfs-tests/tests/functional/Makefile.am
index 396124986..2368b829b 100644
--- a/tests/zfs-tests/tests/functional/Makefile.am
+++ b/tests/zfs-tests/tests/functional/Makefile.am
@@ -28,6 +28,7 @@ SUBDIRS = \
 	hkdf \
 	inheritance \
 	inuse \
+	kstat \
 	large_files \
 	largest_pool \
 	libzfs \
diff --git a/tests/zfs-tests/tests/functional/kstat/Makefile.am b/tests/zfs-tests/tests/functional/kstat/Makefile.am
new file mode 100644
index 000000000..8ad83ec3e
--- /dev/null
+++ b/tests/zfs-tests/tests/functional/kstat/Makefile.am
@@ -0,0 +1,5 @@
+pkgdatadir = $(datadir)/@PACKAGE@/zfs-tests/tests/functional/kstat
+dist_pkgdata_SCRIPTS = \
+	setup.ksh \
+	cleanup.ksh \
+	state.ksh
diff --git a/tests/zfs-tests/tests/functional/kstat/cleanup.ksh b/tests/zfs-tests/tests/functional/kstat/cleanup.ksh
new file mode 100755
index 000000000..8a212ce37
--- /dev/null
+++ b/tests/zfs-tests/tests/functional/kstat/cleanup.ksh
@@ -0,0 +1,28 @@
+#!/bin/ksh -p
+#
+# CDDL HEADER START
+#
+# The contents of this file are subject to the terms of the
+# Common Development and Distribution License (the "License").
+# You may not use this file except in compliance with the License.
+#
+# You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
+# or http://www.opensolaris.org/os/licensing.
+# See the License for the specific language governing permissions
+# and limitations under the License.
+#
+# When distributing Covered Code, include this CDDL HEADER in each
+# file and include the License file at usr/src/OPENSOLARIS.LICENSE.
+# If applicable, add the following below this CDDL HEADER, with the
+# fields enclosed by brackets "[]" replaced with your own identifying
+# information: Portions Copyright [yyyy] [name of copyright owner]
+#
+# CDDL HEADER END
+#
+#
+# Copyright (c) 2018 by Lawrence Livermore National Security, LLC.
+#
+
+. $STF_SUITE/include/libtest.shlib
+
+default_cleanup
diff --git a/tests/zfs-tests/tests/functional/kstat/setup.ksh b/tests/zfs-tests/tests/functional/kstat/setup.ksh
new file mode 100755
index 000000000..57717a096
--- /dev/null
+++ b/tests/zfs-tests/tests/functional/kstat/setup.ksh
@@ -0,0 +1,34 @@
+#!/bin/ksh -p
+#
+# CDDL HEADER START
+#
+# The contents of this file are subject to the terms of the
+# Common Development and Distribution License (the "License").
+# You may not use this file except in compliance with the License.
+#
+# You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
+# or http://www.opensolaris.org/os/licensing.
+# See the License for the specific language governing permissions
+# and limitations under the License.
+#
+# When distributing Covered Code, include this CDDL HEADER in each
+# file and include the License file at usr/src/OPENSOLARIS.LICENSE.
+# If applicable, add the following below this CDDL HEADER, with the
+# fields enclosed by brackets "[]" replaced with your own identifying
+# information: Portions Copyright [yyyy] [name of copyright owner]
+#
+# CDDL HEADER END
+#
+#
+# Copyright (c) 2018 by Lawrence Livermore National Security, LLC.
+#
+
+. $STF_SUITE/include/libtest.shlib
+
+if ! is_linux ; then
+	log_unsupported "/proc/spl/kstat/<pool>/health only supported on Linux"
+fi
+
+default_mirror_setup $DISKS
+
+log_pass
diff --git a/tests/zfs-tests/tests/functional/kstat/state.ksh b/tests/zfs-tests/tests/functional/kstat/state.ksh
new file mode 100755
index 000000000..bf0b6e313
--- /dev/null
+++ b/tests/zfs-tests/tests/functional/kstat/state.ksh
@@ -0,0 +1,144 @@
+#!/bin/ksh -p
+#
+# CDDL HEADER START
+#
+# The contents of this file are subject to the terms of the
+# Common Development and Distribution License (the "License").
+# You may not use this file except in compliance with the License.
+#
+# You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
+# or http://www.opensolaris.org/os/licensing.
+# See the License for the specific language governing permissions
+# and limitations under the License.
+#
+# When distributing Covered Code, include this CDDL HEADER in each
+# file and include the License file at usr/src/OPENSOLARIS.LICENSE.
+# If applicable, add the following below this CDDL HEADER, with the
+# fields enclosed by brackets "[]" replaced with your own identifying
+# information: Portions Copyright [yyyy] [name of copyright owner]
+#
+# CDDL HEADER END
+
+#
+# Copyright (c) 2018 by Lawrence Livermore National Security, LLC.
+#
+
+#
+# DESCRIPTION:
+# Test /proc/spl/kstat/zfs/<pool>/state kstat
+#
+# STRATEGY:
+# 1. Create a mirrored pool
+# 2. Check that pool is ONLINE
+# 3. Fault one disk
+# 4. Check that pool is DEGRADED
+# 5. Create a new pool with a single scsi_debug disk
+# 6. Remove the disk
+# 7. Check that pool is SUSPENDED
+# 8. Add the disk back in
+# 9. Clear errors and destroy the pools
+
+. $STF_SUITE/include/libtest.shlib
+
+verify_runnable "both"
+
+function cleanup
+{
+	# Destroy the scsi_debug pool
+	if [ -n "$TESTPOOL2" ] ; then
+		if  [ -n "$host" ] ; then
+			# Re-enable the disk
+			scan_scsi_hosts $host
+
+			# Device may have changed names after being inserted
+			SDISK=$(get_debug_device)
+			log_must ln $DEV_RDSKDIR/$SDISK $REALDISK
+		fi
+
+		# Restore our working pool image
+		if [ -n "$BACKUP" ] ; then
+			gunzip -c $BACKUP > $REALDISK
+			log_must rm -f $BACKUP
+		fi
+
+		# Our disk is back.  Now we can clear errors and destroy the
+		# pool cleanly.
+		log_must zpool clear $TESTPOOL2
+
+		# Now that the disk is back and errors cleared, wait for our
+		# hung 'zpool scrub' to finish.
+		wait
+
+		destroy_pool $TESTPOOL2
+		log_must rm $REALDISK
+		unload_scsi_debug
+	fi
+}
+
+# Check that our pool state values match what's expected
+#
+# $1: pool name
+# $2: expected state ("ONLINE", "DEGRADED", "SUSPENDED", etc)
+function check_all
+{
+	pool=$1
+	expected=$2
+
+	state1=$(zpool status $pool | awk '/state: /{print $2}');
+	state2=$(zpool list -H -o health $pool)
+	state3=$(cat /proc/spl/kstat/zfs/$pool/state)
+	log_note "Checking $expected = $state1 = $state2 = $state3"
+	if [[ "$expected" == "$state1" &&  "$expected" == "$state2" && \
+	    "$expected" == "$state3" ]] ; then
+		true
+	else
+		false
+	fi
+}
+
+log_onexit cleanup
+
+log_assert "Testing /proc/spl/kstat/zfs/<pool>/state kstat"
+
+# Test that the initial pool is healthy
+check_all $TESTPOOL "ONLINE"
+
+# Fault one of the disks, and check that pool is degraded
+DISK1=$(echo "$DISKS" | awk '{print $2}')
+zpool offline -tf $TESTPOOL $DISK1
+check_all $TESTPOOL "DEGRADED"
+
+# Create a new pool out of a scsi_debug disk
+TESTPOOL2=testpool2
+MINVDEVSIZE_MB=$((MINVDEVSIZE / 1048576))
+load_scsi_debug $MINVDEVSIZE_MB 1 1 1 '512b'
+
+SDISK=$(get_debug_device)
+host=$(get_scsi_host $SDISK)
+
+# Use $REALDISK instead of $SDISK in our pool because $SDISK can change names
+# as we remove/add the disk (i.e. /dev/sdf -> /dev/sdg).
+REALDISK=/dev/kstat-state-realdisk
+log_must [ ! -e $REALDISK ]
+ln $DEV_RDSKDIR/$SDISK $REALDISK
+
+log_must zpool create $TESTPOOL2 $REALDISK
+
+# Backup the contents of the disk image
+BACKUP=/tmp/kstat-state-realdisk.gz
+log_must [ ! -e $BACKUP ]
+gzip -c $REALDISK > $BACKUP
+
+# Yank out the disk from under the pool
+log_must rm $REALDISK
+remove_disk $SDISK
+
+# Run a 'zpool scrub' in the background to suspend the pool.  We run it in the
+# background since the command will hang when the pool gets suspended.  The
+# command will resume and exit after we restore the missing disk later on.
+zpool scrub $TESTPOOL2 &
+sleep 1		# Give the scrub some time to run before we check if it fails
+
+log_must check_all $TESTPOOL2 "SUSPENDED"
+
+log_pass "/proc/spl/kstat/zfs/<pool>/state test successful"