From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mx1.redhat.com ([209.132.183.28]:32868 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751850AbbCYTcj (ORCPT ); Wed, 25 Mar 2015 15:32:39 -0400 Received: from int-mx09.intmail.prod.int.phx2.redhat.com (int-mx09.intmail.prod.int.phx2.redhat.com [10.5.11.22]) by mx1.redhat.com (Postfix) with ESMTPS id 41DE8C3289 for ; Wed, 25 Mar 2015 19:32:39 +0000 (UTC) Date: Wed, 25 Mar 2015 15:32:36 -0400 From: Brian Foster Subject: Re: [PATCH 3/9 RESEND] generic: test hardlink to unlinked file Message-ID: <20150325193236.GD3436@laptop.bfoster> References: <1426846618-23413-1-git-send-email-eguan@redhat.com> <1426846618-23413-4-git-send-email-eguan@redhat.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <1426846618-23413-4-git-send-email-eguan@redhat.com> Sender: fstests-owner@vger.kernel.org To: Eryu Guan Cc: fstests@vger.kernel.org, Eric Sandeen List-ID: On Fri, Mar 20, 2015 at 06:16:52PM +0800, Eryu Guan wrote: > Kernel commit > > aae8a97 fs: Don't allow to create hardlink for deleted file > > disabled hardlink to unlinked file. > > Test the race between link and unlink, which could end up adding link > count to an unlinked file and leading to fs corruption on xfs. > > Test case was originally written by Eric Sandeen. > > Cc: Eric Sandeen > Signed-off-by: Eryu Guan > --- > tests/generic/073 | 99 +++++++++++++++++++++++++++++++++++++++++++++++++++ > tests/generic/073.out | 2 ++ > tests/generic/group | 1 + > 3 files changed, 102 insertions(+) > create mode 100755 tests/generic/073 > create mode 100644 tests/generic/073.out > > diff --git a/tests/generic/073 b/tests/generic/073 > new file mode 100755 > index 0000000..6a5261c > --- /dev/null > +++ b/tests/generic/073 > @@ -0,0 +1,99 @@ > +#! /bin/bash > +# FS QA Test No. 073 > +# > +# Test hardlink to unlinked file. > +# > +# Regression test for commit: > +# aae8a97 fs: Don't allow to create hardlink for deleted file > +# > +#----------------------------------------------------------------------- > +# Copyright (c) 2015 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 -f $tmp.* > +} > + > +# get standard environment, filters and checks > +. ./common/rc > +. ./common/filter > + > +# real QA test starts here > + > +# Modify as appropriate. > +_supported_fs generic > +_supported_os Linux > +_require_test > + > +link_unlink_storm() > +{ > + local src=$1 > + local target=$2 > + > + while true; do > + ln -f $src $target >/dev/null 2>&1 > + rm -f $target >/dev/null 2>&1 > + done > +} This seems slightly different from the original reproducer in that it recreates the same link and unlinks/recreates the $src entry many times as well. That seems Ok, but out of curiousity, is this test confirmed to reproduce the original problem? > + > +rm -f $seqres.full > +nr_cpu=`$here/src/feature -o` > +echo "Silence is golden" > + > +# create, open & unlinked files so unlinked inode list is not empty > +for i in `seq 1 64`; do $nr_cpu > + testfile=$TEST_DIR/$seq.unlinked.$i > + touch $testfile > + tail -f $testfile & > + tail_pids="$tail_pids $!" > + rm -f $testfile I get some file not found errors from 'tail' when I run this test. They go away if I add a sleep after the 'tail -f ... &,' so I suspect the rm -f could be racing with the file removal. That might be unnecessarily slow, however. Perhaps create and tail all of the files in one loop, sleep for a bit, then remove them? > +done > + > +# start link/unlink storm > +src=$TEST_DIR/$seq.target > +touch $testfile touch $src ? > +for i in `seq 1 $nr_cpu`; do > + target=$TEST_DIR/$seq.target.link.$i > + link_unlink_storm $src $target & > + link_pids="$link_pids $!" > +done > + > +# remove & re-create target to race with link/unlink > +while true; do > + rm -f $src > + touch $src > +done & > +sleep 5 > +kill $! >/dev/null 2>&1 > + > +kill $tail_pids $link_pids >/dev/null 2>&1 > +wait $tail_pids $link_pids > + > +# all done, no oops/hang expected, _check_filesystems checks TEST_DEV after test > +status=0 > +exit > diff --git a/tests/generic/073.out b/tests/generic/073.out > new file mode 100644 > index 0000000..d107704 > --- /dev/null > +++ b/tests/generic/073.out > @@ -0,0 +1,2 @@ > +QA output created by 073 > +Silence is golden > diff --git a/tests/generic/group b/tests/generic/group > index c0210d2..ee5b642 100644 > --- a/tests/generic/group > +++ b/tests/generic/group > @@ -75,6 +75,7 @@ > 070 attr udf auto quick stress > 071 auto quick prealloc > 072 auto quick quota > +073 auto metadata quick seems fine as well. Brian > 074 rw udf auto > 075 rw udf auto quick > 076 metadata rw udf auto quick stress > -- > 2.1.0 > > -- > To unsubscribe from this list: send the line "unsubscribe fstests" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html