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 >> > _______________________________________________ xfs mailing list xfs@xxxxxxxxxxx http://oss.sgi.com/mailman/listinfo/xfs
