xfstests: add disk failure simulation test

There are many situations where disk may fail for example
1) brutal usb dongle unplug
2) iscsi (or any other netbdev) failure due to network issues
In this situation filesystem which use this blockdevice is
expected to fail(force RO remount, abort, etc) but whole system
should still be operational. In other words:
1) Kernel should not panic
2) Memory should not leak
3) Data integrity operations (sync,fsync,fdatasync, directio) should fail
   for affected filesystem
4) It should be possible to umount broken filesystem

Later when disk becomes available again we expect(only for journaled filesystems):
5) It will be possible to mount filesystem w/o explicit fsck (in order to catch
   issues like https://patchwork.kernel.org/patch/1983981/)
6) Filesystem should be operational
7) After mount/umount has being done all errors should be fixed so fsck should
   not spot any issues.

This test use fault injection (CONFIG_FAULT_INJECTION=y,
CONFIG_FAIL_MAKE_REQUEST=y and CONFIG_FAULT_INJECTION_DEBUG_FS=y config
options) to force all new IO requests to fail for a given device. Xfs
already has XFS_IOC_GOINGDOWN ioctl which provides similar behavior, but it
is fs specific and it does it in an easy way because it performs freeze_bdev()
before actual shutdown.

Test run fsstress in background and then force disk failure.
Once disk failed it check that (1)-(4) is true.
Then makes disk available again and check that (5)-(7) is also true

BE CAREFUL!! test known to cause memory corruption for XFS
see: https://gist.github.com/dmonakhov/4953045

Signed-off-by: Dmitry Monakhov <dmonakhov@openvz.org>
Reviewed-by: Rich Johnston <rjohnston@sgi.com>
Signed-off-by: Rich Johnston <rjohnston@sgi.com>

5 files changed, 210 insertions, 0 deletions

diff --git a/305 b/305
new file mode 100644
index 00000000..63934611
--- /dev/null
+++ b/305
@@ -0,0 +1,186 @@
+#! /bin/bash
+# FSQA Test No. 305
+#
+# Run fsstress and fio(dio/aio and mmap) and simulate disk failure
+# check filesystem consistency at the end.
+#
+#-----------------------------------------------------------------------
+# (c) 2013 Dmitry Monakhov
+#
+# This program is free software; you can redistribute it and/or
+# modify it under the terms of the GNU General Public License as
+# published by the Free Software Foundation.
+#
+# This program is distributed in the hope that it would be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program; if not, write the Free Software Foundation,
+# Inc.,  51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+#
+#-----------------------------------------------------------------------
+#
+# creator
+owner=dmonakhov@openvz.org
+
+seq=`basename $0`
+echo "QA output created by $seq"
+
+here=`pwd`
+tmp=/tmp/$$
+status=1	# failure is the default!
+
+# get standard environment, filters and checks
+. ./common.rc
+. ./common.filter
+_supported_fs ext3 ext4 xfs btrfs reiserfs
+_supported_os Linux
+_need_to_be_root
+_require_scratch
+_require_fail_make_request
+
+# TODO: Function are common enough to be moved to common.blkdev
+SCRATCH_REAL_DEV=`readlink -f $SCRATCH_DEV`
+SCRATCH_BDEV=`basename $SCRATCH_REAL_DEV`
+
+allow_fail_make_request()
+{
+    echo "Allow global fail_make_request feature"
+    echo 100 > $DEBUGFS_MNT/fail_make_request/probability
+    echo 9999999 > $DEBUGFS_MNT/fail_make_request/times
+    echo 0 >  /sys/kernel/debug/fail_make_request/verbose
+}
+
+disallow_fail_make_request()
+{
+    echo "Disallow global fail_make_request feature"
+    echo 0 > $DEBUGFS_MNT/fail_make_request/probability
+    echo 0 > $DEBUGFS_MNT/fail_make_request/times
+}
+
+start_fail_scratch_dev()
+{
+    echo "Force SCRATCH_DEV device failure"
+    echo " echo 1 > /sys/block/$SCRATCH_BDEV/make-it-fail" >> $here/$seq.full
+    echo 1 > /sys/block/$SCRATCH_BDEV/make-it-fail
+
+}
+
+stop_fail_scratch_dev()
+{
+    echo "Make SCRATCH_DEV device operable again"
+    echo " echo 0 > /sys/block/$SCRATCH_BDEV/make-it-fail" >> $here/$seq.full
+    echo 0 > /sys/block/$SCRATCH_BDEV/make-it-fail
+
+}
+
+_cleanup()
+{
+    poweron_scratch_dev
+    disallow_fail_make_request
+}
+trap "_cleanup; exit \$status" 1 2 3 15
+
+RUN_TIME=$((20+10*$TIME_FACTOR))
+NUM_JOBS=$((4*LOAD_FACTOR))
+BLK_DEV_SIZE=`blockdev --getsz $SCRATCH_DEV`
+FILE_SIZE=$((BLK_DEV_SIZE * 512))
+
+cat >$tmp-$seq.fio <<EOF
+###########
+# $seq test's fio activity
+# Filenames derived from jobsname and jobid like follows:
+# ${JOB_NAME}.${JOB_ID}.${ITERATION_ID}
+[global]
+ioengine=libaio
+bs=4k
+directory=${SCRATCH_MNT}
+filesize=${FILE_SIZE}
+size=9999T
+continue_on_error=write
+ignore_error=EIO,ENOSPC:EIO
+error_dump=0
+
+[stress_dio_aio_activity]
+create_on_open=1
+fallocate=none
+iodepth=128*${LOAD_FACTOR}
+direct=1
+buffered=0
+numjobs=${NUM_JOBS}
+rw=randwrite
+runtime=40+${RUN_TIME}
+time_based
+
+[stress_mmap_activity]
+ioengine=mmap
+create_on_open=0
+fallocate=1
+fdatasync=40960
+filesize=8M
+size=9999T
+numjobs=${NUM_JOBS}
+rw=randwrite
+runtime=40+${RUN_TIME}
+time_based
+
+EOF
+
+_require_fio $tmp-$seq.fio
+
+# Disable all sync operations to get higher load
+FSSTRESS_AVOID="$FSSTRESS_AVOID -ffsync=0 -fsync=0 -ffdatasync=0 -f setattr=1"
+
+_workout()
+{
+	out=$SCRATCH_MNT/fsstress.$$
+	args=`_scale_fsstress_args -p 1 -n999999999 -f setattr=0 $FSSTRESS_AVOID -d $out`
+	echo ""
+	echo "Start fsstress.."
+	echo ""
+	echo "fsstress $args" >> $here/$seq.full
+	$FSSTRESS_PROG $args > /dev/null 2>&1 &
+	fs_pid=$!
+	echo "Start fio.."
+	cat $tmp-$seq.fio >>  $seq.full
+	$FIO_PROG $tmp-$seq.fio >> $here/$seq.full 2>&1 &
+	fio_pid=$!
+
+	# Let's it work for awhile, and force device failure
+	sleep $RUN_TIME
+	start_fail_scratch_dev
+	# After device turns in to failed state filesystem may yet not know about
+	# that so buffered write(2) may succeed, but any integrity operations
+	# such as (sync, fsync, fdatasync, direct-io) should fail.
+	dd if=/dev/zero of=$SCRATCH_MNT/touch_failed_filesystem count=1 bs=4k conv=fsync \
+	    >> $here/$seq.full 2>&1 && \
+	    _fail "failed: still able to perform integrity fsync on $SCRATCH_MNT"
+
+	kill $fs_pid
+	wait $fs_pid
+	wait $fio_pid
+
+	# We expect that broken FS still can be umounted
+	run_check umount $SCRATCH_DEV
+	# Once filesystem was umounted no one is able to write to block device
+	# It is now safe to bring device back to normal state
+	stop_fail_scratch_dev
+
+	# In order to check that filesystem is able to recover journal on mount(2)
+	# perform mount/umount, after that all errors should be fixed
+	run_check _scratch_mount
+	run_check _scratch_unmount
+	_check_scratch_fs
+}
+
+# real QA test starts here
+
+_scratch_mkfs >> $here/$seq.full 2>&1 || _fail "mkfs failed"
+_scratch_mount || _fail "mount failed"
+allow_fail_make_request
+_workout
+status=$?
+disallow_fail_make_request
+exit
diff --git a/305.out b/305.out
new file mode 100644
index 00000000..1e8cfd62
--- /dev/null
+++ b/305.out
@@ -0,0 +1,9 @@
+QA output created by 305
+Allow global fail_make_request feature
+
+Start fsstress..
+
+Start fio..
+Force SCRATCH_DEV device failure
+Make SCRATCH_DEV device operable again
+Disallow global fail_make_request feature
diff --git a/common.config b/common.config
index 951587da..e3f9e452 100644
--- a/common.config
+++ b/common.config
@@ -75,6 +75,7 @@ export BENCH_PASSES=${BENCH_PASSES:=5}
 export XFS_MKFS_OPTIONS=${XFS_MKFS_OPTIONS:=-bsize=4096}
 export TIME_FACTOR=${TIME_FACTOR:=1}
 export LOAD_FACTOR=${LOAD_FACTOR:=1}
+export DEBUGFS_MNT=${DEBUGFS_MNT:="/sys/kernel/debug"}
 
 export PWD=`pwd`
 #export DEBUG=${DEBUG:=...} # arbitrary CFLAGS really.
diff --git a/common.rc b/common.rc
index f5180fe4..ef41cbf6 100644
--- a/common.rc
+++ b/common.rc
@@ -1049,6 +1049,19 @@ _require_sparse_files()
     esac
 }
 
+_require_debugfs()
+{
+    #boot_params always present in debugfs
+    [ -d "$DEBUGFS_MNT/boot_params" ] || _notrun "Debugfs not mounted"
+}
+
+_require_fail_make_request()
+{
+    [ -f "$DEBUGFS_MNT/fail_make_request/probability" ] \
+	|| _notrun "$DEBUGFS_MNT/fail_make_request \
+ not found. Seems that CONFIG_FAIL_MAKE_REQUEST kernel config option not enabled"
+}
+
 # check that a FS on a device is mounted
 # if so, return mount point
 #
diff --git a/group b/group
index da975d3e..c3c41611 100644
--- a/group
+++ b/group
@@ -428,3 +428,4 @@ stress
 302 aio dangerous ioctl rw stress
 303 aio dangerous ioctl rw stress
 304 aio dangerous ioctl rw stress
+305 aio dangerous enospc rw stress