lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  linux-cve-announce  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Date:   Fri, 14 Jul 2017 15:25:29 +0800
From:   Eryu Guan <eguan@...hat.com>
To:     Ernesto A. Fernández 
        <ernesto.mnd.fernandez@...il.com>
Cc:     fstests@...r.kernel.org, Christoph Hellwig <hch@...radead.org>,
        linux-ext4@...r.kernel.org, jfs-discussion@...ts.sourceforge.net,
        reiserfs-devel@...r.kernel.org
Subject: Re: [xfstests PATCH v2] generic: add test of file mode when setfacl
 fails

On Fri, Jul 14, 2017 at 03:59:28AM -0300, Ernesto A. Fernández wrote:
> Check that the group permission bits of a file are not altered when setfacl
> fails. At the time of this patch the test fails for at least ext2, ext4 and
> jfs.
> 
> Also add support for jfs in _scratch_mkfs_sized, which is needed here.
> 
> Note that this test is not meaningful for all filesystems, because some
> will still succeed in setting the acls. This does not mean they don't have
> a bug in how they would handle an error.
> 
> Signed-off-by: Ernesto A. Fernández <ernesto.mnd.fernandez@...il.com>
> ---
> Changes in v2:
>   - Use the scratch dev instead of the test dev so we can make a small fs and
>     fill it faster. As suggested by Christoph Hellwig.
>   - Make jfs work with _scratch_mkfs_sized, needed for the above.
>   - Remove the mention to an issue with reiserfs, since I now believe the
>     problem is on their side.

Thanks for the updated patch! Some minor comments inline.

> 
>  common/rc             |  5 ++-
>  tests/generic/447     | 91 +++++++++++++++++++++++++++++++++++++++++++++++++++
>  tests/generic/447.out |  2 ++
>  tests/generic/group   |  1 +
>  4 files changed, 98 insertions(+), 1 deletion(-)
>  create mode 100755 tests/generic/447
>  create mode 100644 tests/generic/447.out
> 
> diff --git a/common/rc b/common/rc
> index b505365..a2258be 100644
> --- a/common/rc
> +++ b/common/rc
> @@ -1020,7 +1020,7 @@ _scratch_mkfs_sized()
>      xfs)
>  	def_blksz=`echo $MKFS_OPTIONS|sed -rn 's/.*-b ?size= ?+([0-9]+).*/\1/p'`
>  	;;
> -    ext2|ext3|ext4|ext4dev|udf|btrfs|reiser4|ocfs2)
> +    ext2|ext3|ext4|ext4dev|udf|btrfs|reiser4|ocfs2|jfs)
>  	def_blksz=`echo $MKFS_OPTIONS| sed -rn 's/.*-b ?+([0-9]+).*/\1/p'`
>  	;;
>      esac
> @@ -1068,6 +1068,9 @@ _scratch_mkfs_sized()
>  	(( fssize <= 100 * 1024 * 1024 )) && mixed_opt='--mixed'
>  	$MKFS_BTRFS_PROG $MKFS_OPTIONS $mixed_opt -b $fssize $SCRATCH_DEV
>  	;;
> +    jfs)
> +	${MKFS_PROG}.$FSTYP -F $MKFS_OPTIONS -b $blocksize $SCRATCH_DEV $blocks
> +	;;
>      reiser4)
>  	# mkfs.resier4 requires size in KB as input for creating filesystem
>  	$MKFS_REISER4_PROG $MKFS_OPTIONS -y -b $blocksize $SCRATCH_DEV \

Could you please add jfs support to _scratch_mkfs_sized in a separate
patch?

> diff --git a/tests/generic/447 b/tests/generic/447
> new file mode 100755
> index 0000000..ff81355
> --- /dev/null
> +++ b/tests/generic/447
> @@ -0,0 +1,91 @@
> +#! /bin/bash
> +# FS QA Test 447
> +#
> +# Fill the device and set as many extended attributes to a file as
> +# possible. Then call setfacl on it and, if this fails for lack of
> +# space, test that the permissions remain the same.
> +#
> +#-----------------------------------------------------------------------
> +# Copyright (c) 2017 Ernesto A. Fernandez.  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 /
> +	_scratch_unmount

This is not needed, SCRATCH_DEV will be unmounted by next test that
calls _require_scratch.

> +	rm -f $tmp.*
> +}
> +
> +# get standard environment, filters and checks
> +. ./common/rc
> +. ./common/filter
> +. ./common/attr
> +
> +# remove previous $seqres.full before test
> +rm -f $seqres.full
> +
> +# real QA test starts here
> +
> +# Modify as appropriate.
> +_supported_fs generic
> +_supported_os Linux
> +_require_scratch
> +_require_test
> +_require_acls
> +_require_attrs
> +
> +_scratch_mkfs_sized $((50 * 1024 * 1024)) >> $seqres.full 2>&1
> +_scratch_mount

Because this test is trying to consume all space on $SCRATCH_MNT, it'd
be good to exit if _scratch_mount fails, otherwise it's filling rootfs.

_scratch_mount || _fail "mount failed"

> +
> +cd $SCRATCH_MNT
> +TFILE=testfile.$seq

Hmm, fstests prefer:

TFILE=$SCRATCH_MNT/testfile.$seq
touch $TFILE
chmod u_rwx $TFILE
...

instead of cd to $SCRATCH_MNT.

> +
> +# Create the test file and choose its permissions
> +rm -f $TFILE

$SCRATCH_DEV was just created, no need to remove $TFILE here.

> +touch $TFILE
> +chmod u+rwx $TFILE
> +chmod go-rwx $TFILE
> +
> +# Try to run out of space so setfacl will fail
> +yes xxxxxxxxxx &> $TFILE

How about just writing 50M to $TFILE?

$XFS_IO_PROG -fc "pwrite 0 50m" $TFILE >>$seqres.full 2>&1

> +i=1
> +while setfattr -n user.$i $TFILE &> /dev/null; do
> +	((++i))
> +done
> +
> +if setfacl -m m:r $TFILE &> /dev/null; then
> +	# setfacl succeeded, so the test was meaningless
> +	# The filesystem might still have an issue
> +	status=0
> +	echo "-rwx------"
> +	exit
> +fi
> +
> +# Since setfacl failed, the permissions should not have changed
> +stat -c %A $TFILE
> +
> +status=0
> +exit
> diff --git a/tests/generic/447.out b/tests/generic/447.out
> new file mode 100644
> index 0000000..adec877
> --- /dev/null
> +++ b/tests/generic/447.out
> @@ -0,0 +1,2 @@
> +QA output created by 447
> +-rwx------
> diff --git a/tests/generic/group b/tests/generic/group
> index 8c1e21a..f11d798 100644
> --- a/tests/generic/group
> +++ b/tests/generic/group
> @@ -449,3 +449,4 @@
>  444 auto quick acl
>  445 auto quick rw
>  446 auto quick rw dangerous
> +447 acl

Should be in auto and quick group too. Basically auto group means this
test is a well defined regression test and the run time is less than
300s, quick group is a subset of auto, and the run time is usually
within 30s.

Thanks,
Eryu

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ