From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from cuda.sgi.com (cuda3.sgi.com [192.48.176.15]) by oss.sgi.com (8.14.3/8.14.3/SuSE Linux 0.8) with ESMTP id o88Iow98055060 for ; Wed, 8 Sep 2010 13:50:59 -0500 Received: from mail.sandeen.net (localhost [127.0.0.1]) by cuda.sgi.com (Spam Firewall) with ESMTP id E658317BCE4E for ; Wed, 8 Sep 2010 11:51:42 -0700 (PDT) Received: from mail.sandeen.net (64-131-60-146.usfamily.net [64.131.60.146]) by cuda.sgi.com with ESMTP id NNjbP88yVJfTtGHX for ; Wed, 08 Sep 2010 11:51:42 -0700 (PDT) Message-ID: <4C87DB3D.4030709@sandeen.net> Date: Wed, 08 Sep 2010 13:51:41 -0500 From: Eric Sandeen MIME-Version: 1.0 Subject: Re: [PATCH] Test to ensure that the EOFBLOCK_FL gets set/unset correctly. References: <1282941224-5805-1-git-send-email-alal@google.com> <20100827233216.GJ705@dastard> <4C785254.2020708@sandeen.net> <20100828001703.GK705@dastard> <4C785710.9070405@sandeen.net> <4C868314.9090503@sandeen.net> In-Reply-To: List-Id: XFS Filesystem from SGI List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Sender: xfs-bounces@oss.sgi.com Errors-To: xfs-bounces@oss.sgi.com To: alal@google.com Cc: xfs@oss.sgi.com 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 > --------------------------------------------------------------------------------------- > Updated patch: > --------------------------------------------------------------------------------------- > From e6906071ab6c0ad38d3ee0a463b5c7944e71fd00 Mon Sep 17 00:00:00 2001 > From: Akshay Lal > 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 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 +# 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@google.com > + > +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@oss.sgi.com http://oss.sgi.com/mailman/listinfo/xfs