Akshay Lal wrote:
> Thanks Eric, as we move fwd, these patch will be better.
>
No problem, I appreciate the testcase... a lot! :)
-Eric
> ---
> Cheers!
> Akshay Lal
>
>
>
> On Wed, Sep 8, 2010 at 11:51 AM, Eric Sandeen <sandeen@xxxxxxxxxxx> wrote:
>> Akshay Lal wrote:
>>> There was a missing fi at the the end of the _check_ext4_eof_flag method.
>>>
>>> Also corrected some indentation.
>> Thanks, I've committed this version.
>>
>> In the future, keeping patch resubmissions tidy is helpful, i.e.
>>
>> Subject: [PATCH Vx] subsystem: description
>>
>> patch description
>>
>> Signed-off-by:
>> ---
>> other information that won't go into the commit
>>
>> patch itself
>>
>> ... without the other cruft, forwarded snippets, etc :)
>>
>> see also Documentation/SubmittingPatches
>> in the kernel tree.
>>
>> Thanks,
>>
>> -Eric
>>
>>> Signed-off-by: Akshay Lal <alal@xxxxxxxxxx>
>>> ---------------------------------------------------------------------------------------
>>> Updated patch:
>>> ---------------------------------------------------------------------------------------
>>> From e6906071ab6c0ad38d3ee0a463b5c7944e71fd00 Mon Sep 17 00:00:00 2001
>>> From: Akshay Lal <alal@xxxxxxxxxx>
>>> Date: Tue, 7 Sep 2010 12:54:33 -0700
>>> Subject: [PATCH] Test to ensure that the EOFBLOCK_FL gets set/unset
>>> correctly.
>>>
>>> As found by Theodore Ts'o:
>>> If a 128K file is falloc'ed using the KEEP_SIZE flag, and then
>>> write exactly 128K, the EOFBLOCK_FL doesn't get cleared correctly.
>>> This is bad since it forces e2fsck to complain about that inode.
>>> If you have a large number of inodes that are written with fallocate
>>> using KEEP_SIZE, and then fill them up to their expected size,
>>> e2fsck will potentially complain about a _huge_ number of inodes.
>>> This would also cause a huge increase in the time taken by e2fsck
>>> to complete its check.
>>>
>>> Test scenarios covered:
>>> 1. Fallocating X bytes and writing Y (Y<X) (buffered and direct io)
>>> 2. Fallocating X bytes and writing Y (Y=X) (buffered and direct io)
>>> 3. Fallocating X bytes and writing Y (Y>X) (buffered and direct io)
>>>
>>> These test cases exercise the normal and edge case conditions using
>>> falloc (and KEEP_SIZE).
>>>
>>> Ref: http://thread.gmane.org/gmane.comp.file-systems.ext4/20682
>>> ---
>>> 243 | 178
>>> +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
>>> 243.out | 13 +++++
>>> group | 1 +
>>> 3 files changed, 192 insertions(+), 0 deletions(-)
>>> create mode 100644 243
>>> create mode 100644 243.out
>>>
>>> diff --git a/243 b/243
>>> new file mode 100644
>>> index 0000000..8b2e647
>>> --- /dev/null
>>> +++ b/243
>>> @@ -0,0 +1,178 @@
>>> +#! /bin/bash
>>> +# FS QA Test No. 243
>>> +#
>>> +# Test to ensure that the EOFBLOCK_FL gets set/unset correctly.
>>> +#
>>> +# As found by Theodore Ts'o:
>>> +# If a 128K file is falloc'ed using the KEEP_SIZE flag, and then
>>> +# write exactly 128K, the EOFBLOCK_FL doesn't get cleared correctly.
>>> +# This is bad since it forces e2fsck to complain about that inode.
>>> +# If you have a large number of inodes that are written with fallocate
>>> +# using KEEP_SIZE, and then fill them up to their expected size,
>>> +# e2fsck will potentially complain about a _huge_ number of inodes.
>>> +# This would also cause a huge increase in the time taken by e2fsck
>>> +# to complete its check.
>>> +#
>>> +# Test scenarios covered:
>>> +# 1. Fallocating X bytes and writing Y (Y<X) (buffered and direct io)
>>> +# 2. Fallocating X bytes and writing Y (Y=X) (buffered and direct io)
>>> +# 3. Fallocating X bytes and writing Y (Y>X) (buffered and direct io)
>>> +#
>>> +# These test cases exercise the normal and edge case conditions using
>>> +# falloc (and KEEP_SIZE).
>>> +#
>>> +# Ref: http://thread.gmane.org/gmane.comp.file-systems.ext4/20682
>>> +#
>>> +#-----------------------------------------------------------------------
>>> +# Copyright (c) 2010 Google, 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
>>> +#-----------------------------------------------------------------------
>>> +#
>>> +# creator
>>> +owner=alal@xxxxxxxxxx
>>> +
>>> +seq=`basename $0`
>>> +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
>>> +
>>> +# Test specific macros.
>>> +BIT_NOT_SET=0 # inode flag - 0x400000 bit is not set.
>>> +BIT_SET=1 # inode flag - 0x400000 bit is set.
>>> +
>>> +# Generic test cleanup function.
>>> +_cleanup()
>>> +{
>>> + cd /
>>> + rm -f $tmp.*
>>> +}
>>> +
>>> +# Ext4 uses the EOFBLOCKS_FL bit when fallocating blocks with KEEP_SIZE
>>> +# enabled. The only time this bit should be set is when extending the
>>> allocated
>>> +# blocks further than what the i_size represents. In the situations
>>> wherein the
>>> +# i_size covers all allocated blocks, this bit should be cleared.
>>> +
>>> +# Checks the state of the sample file in the filesystem and returns whether
>>> +# the inode flag 0x400000 is set or not.
>>> +_check_ext4_eof_flag()
>>> +{
>>> + # Check whether EOFBLOCK_FL is set.
>>> + # For ext4 filesystems: use debugfs to check if EOFBLOCKS_FL is set.
>>> + # Other filesystems: do nothing. The default fsck at the end of the test
>>> + # should catch any potential errors.
>>> + if [ "${FSTYP}" == "ext4" ]; then
>>> + bit_set=1
>>> +
>>> + # Unmount the ${TEST_DEV}
>>> + umount ${TEST_DEV}
>>> +
>>> + # Run debugfs to gather file_parameters - specifically iflags.
>>> + file_params=`debugfs ${TEST_DEV} -R "stat ${1}" 2>&1 | grep -e Flags:`
>>> + iflags=${file_params#*Flags: }
>>> +
>>> + # Ensure that the iflags value was parsed correctly.
>>> + if [ -z ${iflags} ]; then
>>> + echo "iFlags value was not parsed successfully." >> $seq.full
>>> + status=1
>>> + exit ${status}
>>> + fi
>>> +
>>> + # Check if EOFBLOCKS_FL is set.
>>> + if ((${iflags} & 0x400000)); then
>>> + echo "EOFBLOCK_FL bit is set." >> $seq.full
>>> + bit_set=1
>>> + else
>>> + echo "EOFBLOCK_FL bit is not set." >> $seq.full
>>> + bit_set=0
>>> + fi
>>> +
>>> + # Check current bit state to expected value.
>>> + if [ ${bit_set} -ne ${2} ]; then
>>> + echo "Error: Current bit state incorrect." >> $seq.full
>>> + status=1
>>> + exit ${status}
>>> + fi
>>> +
>>> + # Mount the ${TEST_DEV}
>>> + mount ${TEST_DEV} -t ${FSTYP} ${TEST_DIR}
>>> + fi
>>> +}
>>> +
>>> +# Get standard environment, filters and checks.
>>> +. ./common.rc
>>> +. ./common.filter
>>> +
>>> +# Prerequisites for the test run.
>>> +_supported_fs ext4 xfs btrfs gfs2
>>> +_supported_os Linux
>>> +_require_xfs_io_falloc
>>> +
>>> +# Real QA test starts here.
>>> +rm -f $seq.full
>>> +
>>> +# Begin test cases.
>>> +echo "Test 1: Fallocate 40960 bytes and write 4096 bytes (buffered io)." \
>>> + >> $seq.full
>>> +${XFS_IO_PROG} -F -f \
>>> + -c 'falloc -k 0 40960' \
>>> + -c 'pwrite 0 4096' \
>>> + ${TEST_DIR}/test_1 | _filter_xfs_io_unique
>>> +_check_ext4_eof_flag test_1 ${BIT_SET}
>>> +
>>> +echo "Test 2: Fallocate 40960 bytes and write 4096 bytes (direct io)." \
>>> + >> $seq.full
>>> +${XFS_IO_PROG} -F -f -d \
>>> + -c 'falloc -k 0 40960' \
>>> + -c 'pwrite 0 4096' \
>>> + ${TEST_DIR}/test_2 | _filter_xfs_io_unique
>>> +_check_ext4_eof_flag test_2 ${BIT_SET}
>>> +
>>> +echo "Test 3: Fallocate 40960 bytes and write 40960 bytes (buffered io)." \
>>> + >> $seq.full
>>> +${XFS_IO_PROG} -F -f \
>>> + -c 'falloc -k 0 40960' \
>>> + -c 'pwrite 0 40960' \
>>> + ${TEST_DIR}/test_3 | _filter_xfs_io_unique
>>> +_check_ext4_eof_flag test_3 ${BIT_NOT_SET}
>>> +
>>> +echo "Test 4: Fallocate 40960 bytes and write 40960 bytes (direct io)." \
>>> + >> $seq.full
>>> +${XFS_IO_PROG} -F -f -d \
>>> + -c 'falloc -k 0 40960' \
>>> + -c 'pwrite 0 40960' \
>>> + ${TEST_DIR}/test_4 | _filter_xfs_io_unique
>>> +_check_ext4_eof_flag test_4 ${BIT_NOT_SET}
>>> +
>>> +echo "Test 5: Fallocate 128k, seek 256k and write 4k block (buffered io)."
>>> \
>>> + >> $seq.full
>>> +${XFS_IO_PROG} -F -f \
>>> + -c 'falloc -k 0 128k' \
>>> + -c 'pwrite 256k 4k' \
>>> + ${TEST_DIR}/test_5 | _filter_xfs_io_unique
>>> +_check_ext4_eof_flag test_5 ${BIT_NOT_SET}
>>> +
>>> +echo "Test 6: Fallocate 128k, seek to 256k and write a 4k block (direct
>>> io)." \
>>> + >> $seq.full
>>> +${XFS_IO_PROG} -F -f -d \
>>> + -c 'falloc -k 0 128k' \
>>> + -c 'pwrite 256k 4k' \
>>> + ${TEST_DIR}/test_6 | _filter_xfs_io_unique
>>> +_check_ext4_eof_flag test_6 ${BIT_NOT_SET}
>>> +
>>> +status=0
>>> +exit ${status}
>>> diff --git a/243.out b/243.out
>>> new file mode 100644
>>> index 0000000..290a005
>>> --- /dev/null
>>> +++ b/243.out
>>> @@ -0,0 +1,13 @@
>>> +QA output created by 243
>>> +wrote 4096/4096 bytes at offset 0
>>> +XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
>>> +wrote 4096/4096 bytes at offset 0
>>> +XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
>>> +wrote 40960/40960 bytes at offset 0
>>> +XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
>>> +wrote 40960/40960 bytes at offset 0
>>> +XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
>>> +wrote 4096/4096 bytes at offset 262144
>>> +XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
>>> +wrote 4096/4096 bytes at offset 262144
>>> +XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
>>> diff --git a/group b/group
>>> index ff16bb3..e6dab13 100644
>>> --- a/group
>>> +++ b/group
>>> @@ -356,3 +356,4 @@ deprecated
>>> 240 auto aio quick rw
>>> 241 auto
>>> 242 auto quick prealloc
>>> +243 auto quick prealloc
>>
>
|