From: Eric Biggers <ebiggers@google.com>
To: fstests@vger.kernel.org
Cc: Theodore Ts'o <tytso@mit.edu>, Jaegeuk Kim <jaegeuk@kernel.org>,
Richard Weinberger <richard@nod.at>,
David Gstir <david@sigma-star.at>,
Michael Halcrow <mhalcrow@google.com>,
Eric Biggers <ebiggers@google.com>
Subject: [PATCH v2 5/5] generic: test for weaknesses in filesystem encryption
Date: Mon, 28 Nov 2016 14:16:42 -0800 [thread overview]
Message-ID: <1480371402-12204-6-git-send-email-ebiggers@google.com> (raw)
In-Reply-To: <1480371402-12204-1-git-send-email-ebiggers@google.com>
Add an xfstest which can detect some basic crypto mistakes that would
reduce the confidentiality guarantee provided by filesystem encryption.
Signed-off-by: Eric Biggers <ebiggers@google.com>
---
common/config | 1 +
tests/generic/403 | 155 ++++++++++++++++++++++++++++++++++++++++++++++++++
tests/generic/403.out | 3 +
tests/generic/group | 1 +
4 files changed, 160 insertions(+)
create mode 100755 tests/generic/403
create mode 100644 tests/generic/403.out
diff --git a/common/config b/common/config
index 3727ec0..6cce7ce 100644
--- a/common/config
+++ b/common/config
@@ -203,6 +203,7 @@ export UUIDGEN_PROG="`set_prog_path uuidgen`"
export GETRICHACL_PROG="`set_prog_path getrichacl`"
export SETRICHACL_PROG="`set_prog_path setrichacl`"
export KEYCTL_PROG="`set_prog_path keyctl`"
+export XZ_PROG="`set_prog_path xz`"
# use 'udevadm settle' or 'udevsettle' to wait for lv to be settled.
# newer systems have udevadm command but older systems like RHEL5 don't.
diff --git a/tests/generic/403 b/tests/generic/403
new file mode 100755
index 0000000..0ca3647
--- /dev/null
+++ b/tests/generic/403
@@ -0,0 +1,155 @@
+#! /bin/bash
+# FS QA Test generic/403
+#
+# Check for weaknesses in filesystem encryption involving the same ciphertext
+# being repeated. For file contents, we fill a small filesystem with large
+# files of 0's and verify the filesystem is incompressible. For filenames, we
+# create an identical symlink in two different directories and verify the
+# ciphertext filenames and symlink targets are different.
+#
+# This test can detect some basic cryptographic mistakes such as nonce reuse
+# (across files), initialization vector reuse (across blocks), or data somehow
+# being left in plaintext by accident. For example, it detects the
+# initialization vector reuse bug fixed in commit 02fc59a0d28f ("f2fs/crypto:
+# fix xts_tweak initialization").
+#
+#-----------------------------------------------------------------------
+# Copyright (c) 2016 Google, Inc. All Rights Reserved.
+#
+# Author: Eric Biggers <ebiggers@google.com>
+#
+# 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
+. ./common/encrypt
+
+# remove previous $seqres.full before test
+rm -f $seqres.full
+
+# real QA test starts here
+_supported_fs ext4 f2fs
+_supported_os Linux
+_require_xfs_io_command "set_encpolicy"
+_require_scratch
+_require_command "$XZ_PROG" xz
+_require_command "$KEYCTL_PROG" keyctl
+_require_encryption
+
+# Set up a small (32 MB) filesystem containing an encrypted directory.
+
+fs_size=$((32 * 1024 * 1024))
+MKFS_OPTIONS="$MKFS_OPTIONS -O encrypt" \
+ _scratch_mkfs_sized $fs_size &>> $seqres.full
+_scratch_mount
+
+keydesc=$(_generate_encryption_key)
+mkdir $SCRATCH_MNT/encrypted_dir
+$XFS_IO_PROG -c "set_encpolicy $keydesc" $SCRATCH_MNT/encrypted_dir
+
+# Create the "same" symlink in two different directories.
+# Later we'll check both the name and target of the symlink.
+mkdir $SCRATCH_MNT/encrypted_dir/subdir1
+mkdir $SCRATCH_MNT/encrypted_dir/subdir2
+ln -s symlink_target $SCRATCH_MNT/encrypted_dir/subdir1/symlink
+ln -s symlink_target $SCRATCH_MNT/encrypted_dir/subdir2/symlink
+
+#
+# Write files of 1 MB of all the same byte until we hit ENOSPC. Note that we
+# must not create sparse files, since the contents of sparse files are not
+# stored on-disk. Also, we create multiple files rather than one big file
+# because we want to test for reuse of per-file keys.
+#
+total_file_size=0
+i=1
+while true; do
+ file=$SCRATCH_MNT/encrypted_dir/file$i
+ if ! xfs_io -f $file -c 'pwrite 0 1M' &> $tmp.out; then
+ if ! grep -q 'No space left on device' $tmp.out; then
+ echo "FAIL: unexpected pwrite failure"
+ cat $tmp.out
+ elif [ -e $file ]; then
+ total_file_size=$((total_file_size + $(stat -c %s $file)))
+ fi
+ break
+ fi
+ total_file_size=$((total_file_size + $(stat -c %s $file)))
+ i=$((i + 1))
+ if [ $i -ge 33 ]; then
+ echo "FAIL: filesystem never filled up!"
+ break
+ fi
+done
+
+# We shouldn't have been able to write more data than we had space for.
+if (( $total_file_size > $fs_size )); then
+ echo "FAIL: wrote $total_file_size bytes but should have only" \
+ "had space for $fs_size bytes at most"
+fi
+
+#
+# Unmount the filesystem and compute its compressed size. It must be no smaller
+# than the amount of data that was written; otherwise there was a compromise in
+# the confidentiality of the data. False positives should not be possible
+# because filesystem metadata will also contribute to the compressed size.
+#
+# Note: it's important to use a strong compressor such as xz which can detect
+# redundancy across the whole filesystem. For a 32 MB filesystem, xz -8 is
+# sufficient because it uses a 32 MB sliding window for compression. At this
+# setting, compression will take 370 MB of memory as documented in xz(1).
+#
+_unlink_encryption_key $keydesc
+_scratch_unmount
+fs_compressed_size=$(head -c $fs_size $SCRATCH_DEV | xz -8 | wc -c)
+
+if (( $fs_compressed_size < $total_file_size )); then
+ echo "FAIL: filesystem was compressible" \
+ "($total_file_size bytes => $fs_compressed_size bytes)"
+else
+ echo "PASS: ciphertexts were not repeated for contents"
+fi
+
+# Verify that encrypted filenames and symlink targets were not reused. Note
+# that since the ciphertexts should be unpredictable, we cannot simply include
+# the expected names in the expected output file.
+_scratch_mount
+find $SCRATCH_MNT/encrypted_dir -type l | wc -l
+link1=$(find $SCRATCH_MNT/encrypted_dir -type l | head -1)
+link2=$(find $SCRATCH_MNT/encrypted_dir -type l | tail -1)
+[ $(basename $link1) = $(basename $link2) ] && \
+ echo "Encrypted filenames were reused!"
+[ $(readlink $link1) = $(readlink $link2) ] && \
+ echo "Encrypted symlink targets were reused!"
+
+# success, all done
+status=0
+exit
diff --git a/tests/generic/403.out b/tests/generic/403.out
new file mode 100644
index 0000000..05c1eaa
--- /dev/null
+++ b/tests/generic/403.out
@@ -0,0 +1,3 @@
+QA output created by 403
+PASS: ciphertexts were not repeated for contents
+2
diff --git a/tests/generic/group b/tests/generic/group
index 7af053d..eb9361e 100644
--- a/tests/generic/group
+++ b/tests/generic/group
@@ -397,3 +397,4 @@
400 auto quick encrypt
401 auto quick encrypt
402 auto quick encrypt
+403 auto encrypt
--
2.8.0.rc3.226.g39d4020
next prev parent reply other threads:[~2016-11-28 22:17 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-11-28 22:16 [PATCH v2 0/5] Add filesystem-level encryption tests Eric Biggers
2016-11-28 22:16 ` [PATCH v2 1/5] generic: add utilities for testing filesystem encryption Eric Biggers
2016-11-28 22:16 ` [PATCH v2 2/5] generic: test setting and getting encryption policies Eric Biggers
2016-11-28 22:16 ` [PATCH v2 3/5] generic: test validation of encryption policy structure Eric Biggers
2016-11-28 22:16 ` [PATCH v2 4/5] generic: test encrypted file access Eric Biggers
2016-11-29 20:52 ` Richard Weinberger
2016-12-01 0:30 ` Eric Biggers
2016-11-28 22:16 ` Eric Biggers [this message]
2016-11-28 22:25 ` [PATCH v2 0/5] Add filesystem-level encryption tests Eric Biggers
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1480371402-12204-6-git-send-email-ebiggers@google.com \
--to=ebiggers@google.com \
--cc=david@sigma-star.at \
--cc=fstests@vger.kernel.org \
--cc=jaegeuk@kernel.org \
--cc=mhalcrow@google.com \
--cc=richard@nod.at \
--cc=tytso@mit.edu \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox