Skip to contentSkip to conversation

xfstests generic/320: heavy rm workload test

PATCH RESEND Superseded Eryu Guan 2 participants 3 messages Nov 12, 2013

v2 of 3 Superseded Latest is v3 →

lore
Share & tools
lore

Eryu Guan

author
#0 lore raw

PATCH RESEND v2 1/1 xfstests generic/320: heavy rm workload test

This test is based on generic/273, a regression test for commit

9a3a5da xfs: check for stale inode before acquiring iflock on push

On unpatched kernel, rm processes would hang.

---
v2: Address Dave's comments
- don't check return value of mkdir
- don't cd to origin dir
- don't calculate fs size, already have fs size
- use $XFS_IO_PROG instead of dd
- remove leading "_" from local vars
- call _supported_fs etc. functions before any local functions
- set status=0 before exit

 tests/generic/320     | 121 ++++++++++++++++++++++++++++++++++++++++++++++++++
 tests/generic/320.out |   2 +
 tests/generic/group   |   2 +-
 3 files changed, 124 insertions(+), 1 deletion(-)
 create mode 100755 tests/generic/320
 create mode 100644 tests/generic/320.out
tests/generic/320 new file +121 -0
new file mode 100755
index 0000000..09c533c
@@ -0,0 +1,121 @@
+#! /bin/bash
+# FS QA Test No. generic/320
+#
+# heavy rm workload
+#
+# Regression test for commit:
+# 9a3a5da xfs: check for stale inode before acquiring iflock on push
+#
+# Based on generic/273
+#
+#-----------------------------------------------------------------------
+# Copyright (c) 2011-2012 Fujitsu, Inc.  All Rights Reserved.
+# Copyright (c) 2013 Red Hat, Inc.  All Rights Reserved.
+#
+# 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
+#
+#-----------------------------------------------------------------------
+#
+
+seq=`basename $0`
+seqres=$RESULT_DIR/$seq
+echo "QA output created by $seq"
+
+here=`pwd`
+tmp=/tmp/$$
+status=1	# failure is the default!
+trap "_cleanup; exit \$status" 0 1 2 3 15
+
+_cleanup()
+{
+	cd /
+	rm -rf $tmp.*
+}
+
+# get standard environment, filters and checks
+. ./common/rc
+. ./common/filter
+
+# real QA test starts here
+_supported_fs generic
+_supported_os IRIX Linux
+_require_scratch
+
+threads=50
+count=2
+fs_size=$((2 * 1024 * 1024 * 1024))
+ORIGIN=$SCRATCH_MNT/origin
+
+threads_set()
+{
+	cpu_num=`grep -c processor /proc/cpuinfo`
+	threads=$(($cpu_num * 50))
+	if [ $threads -gt 200 ]
+	then
+		threads=200
+	fi
+}
+
+file_create()
+{
+	i=0
+	mkdir $ORIGIN
+
+	disksize=$(($fs_size / 3))
+	num=$(($disksize / $count / $threads / 4096))
+	while [ $i -lt $num ]; do
+		$XFS_IO_PROG -f -c "pwrite 0 $((4096*count))" $ORIGIN/file_$i >>$seqres.full 2>&1
+		i=$(($i + 1))
+	done
+}
+
+worker()
+{
+	suffix=$1
+
+	mkdir $SCRATCH_MNT/sub_$suffix
+
+	cp -r $ORIGIN/* $SCRATCH_MNT/sub_$suffix >>$seqres.full 2>&1
+	rm -rf $SCRATCH_MNT/sub_$suffix
+}
+
+do_workload()
+{
+	pids=""
+	loop=1
+
+	threads_set
+	file_create
+
+	while [ $loop -lt $threads ]; do
+		worker $loop &
+		pids="$pids $!"
+		loop=$(($loop + 1))
+	done
+
+	wait $pids
+}
+
+echo "Silence is golden"
+
+rm -f $seqres.full
+
+_scratch_mkfs_sized $fs_size >>$seqres.full 2>&1
+_scratch_mount >>$seqres.full 2>&1
+
+do_workload
+
+_check_scratch_fs
+status=0
+exit
tests/generic/320.out new file +2 -0
new file mode 100644
index 0000000..1c4165d
@@ -0,0 +1,2 @@
+QA output created by 320
+Silence is golden
tests/generic/group +1 -1
index 3488c42..40da6dc 100644
@@ -122,4 +122,4 @@
 317 auto metadata quick
 318 acl attr auto quick
 319 acl auto quick
-
+320 auto rw
Signature · 6 lines
1.8.3.1

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs
#1 lore raw
Quoted 5 lines · Eryu Guan

This test is based on generic/273, a regression test for commit

9a3a5da xfs: check for stale inode before acquiring iflock on push

On unpatched kernel, rm processes would hang.

....

tests/generic/320 line 62

Quoted 8 lines
tests/generic/320 new file +8 -0
+threads=50
+count=2
+fs_size=$((2 * 1024 * 1024 * 1024))
+ORIGIN=$SCRATCH_MNT/origin
+
+threads_set()
+{
+	cpu_num=`grep -c processor /proc/cpuinfo`

Please us src/feature for this. See commit:

2dcf4a5 xfstests: src/feature.c: print a number of online CPUs

tests/generic/320 line 68

Quoted 6 lines
tests/generic/320 new file +6 -0
+	threads=$(($cpu_num * 50))
+	if [ $threads -gt 200 ]
+	then
+		threads=200
+	fi
+}

Didn't we go through this before? Or was that the review that lead to the above commit? i.e. to use $LOAD_FACTOR to scale the workload, not the number of CPUs....

tests/generic/320 line 78

Quoted 10 lines
tests/generic/320 new file +10 -0
+
+file_create()
+{
+	i=0
+	mkdir $ORIGIN
+
+	disksize=$(($fs_size / 3))
+	num=$(($disksize / $count / $threads / 4096))
+	while [ $i -lt $num ]; do
+		$XFS_IO_PROG -f -c "pwrite 0 $((4096*count))" $ORIGIN/file_$i >>$seqres.full 2>&1

Line too long.

Cheers,

Dave.

Signature · 7 lines
Dave Chinner
[email protected]

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs
author
#2 lore raw
Quoted 20 lines · Dave Chinner
Quoted 5 lines · Eryu Guan

This test is based on generic/273, a regression test for commit

9a3a5da xfs: check for stale inode before acquiring iflock on push

On unpatched kernel, rm processes would hang.

....

tests/generic/320 line 62

Quoted 8 lines
tests/generic/320 new file +8 -0
+threads=50
+count=2
+fs_size=$((2 * 1024 * 1024 * 1024))
+ORIGIN=$SCRATCH_MNT/origin
+
+threads_set()
+{
+	cpu_num=`grep -c processor /proc/cpuinfo`

Please us src/feature for this. See commit:

2dcf4a5 xfstests: src/feature.c: print a number of online CPUs

Quoted 10 lines

tests/generic/320 line 68

Quoted 6 lines
tests/generic/320 new file +6 -0
+	threads=$(($cpu_num * 50))
+	if [ $threads -gt 200 ]
+	then
+		threads=200
+	fi
+}

Didn't we go through this before? Or was that the review that lead to the above commit? i.e. to use $LOAD_FACTOR to scale the workload, not the number of CPUs....

Sorry, I may miss that part of review, will revise the patch.

Quoted 13 lines

tests/generic/320 line 78

Quoted 11 lines
tests/generic/320 new file +10 -0
+
+file_create()
+{
+	i=0
+	mkdir $ORIGIN
+
+	disksize=$(($fs_size / 3))
+	num=$(($disksize / $count / $threads / 4096))
+	while [ $i -lt $num ]; do
+		$XFS_IO_PROG -f -c "pwrite 0 $((4096*count))" $ORIGIN/file_$i >>$seqres.full 2>&1

Line too long.

Ok, will update.

Thanks for the review!

Eryu

Quoted 7 lines

Cheers,

Dave.

Signature · 2 lines
Dave Chinner
[email protected]

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs