Fix the zfsd_fault_001_pos test

Due to a regression in ZFS, a pool's status may not change to DEGRADED
promptly after a vdev becomes faulted.  This test is primarily concerned
with vdev status, not pool status, so work around the bug by doing a
"zpool reopen".  Otherwise the test would time out waiting for a status
change that never came.

MFC to stable/14 only.

https://github.com/openzfs/zfs/issues/16245

MFC after:	2 weeks
Sponsored by:	ConnectWise

tests/sys/cddl/zfs/tests/zfsd/zfsd_fault_001_pos.ksh

@@ -78,6 +78,10 @@ for type in "raidz" "mirror"; do
 		$DD if=/dev/zero bs=128k count=1 >> \
 			/$TESTPOOL/$TESTFS/$TESTFILE 2> /dev/null
 		$FSYNC /$TESTPOOL/$TESTFS/$TESTFILE
+		# Due to a bug outside of zfsd, it may be necessary to reopen
+		# the pool before it will become DEGRADED.
+		# https://github.com/openzfs/zfs/issues/16245
+		$ZPOOL reopen $TESTPOOL
 		# Check to see if the pool is faulted yet
 		$ZPOOL status $TESTPOOL | grep -q 'state: DEGRADED'
 		if [ $? == 0 ]