1
0
mirror of https://git.code.sf.net/p/linux-ima/ima-evm-utils synced 2025-04-28 06:33:36 +02:00

Compare commits

..

64 Commits
v1.4 ... master

Author SHA1 Message Date
Mimi Zohar
1803accc3f Release version 1.5
New to this release is CI support for testing new kernel integrity
features not yet upstreamed and bugfixes, or functionality not enabled
by distros in a User Mode Linux (UML) environment.  Testing in a UML
environment also allows saving CI build artifacts, such as private
keys, needed for creating and loading public keys onto the trusted
kernel keyrings.  These public keys may be used for code - file data
and metadata - signature verification.

See the NEWS file for a short summary of changes and the git history
for details.

Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2023-03-06 07:40:07 -05:00
Mimi Zohar
411ff0a720 tests: fix gen-keys.sh to generate sha256 certificates
On systems with OpenSSL sha1 disabled, the sign-verify.test fails:

- openssl dgst   -sha1 sha1.txt
- openssl dgst   -sha1 -sign test-rsa1024.key -hex sha1.txt
Error setting context
804BD5CF787F0000:error:03000098:digital envelope routines:do_sigver_init:invalid digest:crypto/evp/m_sigver.c:343:
sha1 (test-rsa1024.key) test is skipped (openssl is unable to sign)

Instead of enabling sha1 support on these systems by setting the environment
variable OPENSSL_ENABLE_SHA1_SIGNATURES, generate a sha256 certificate.

Reported-by: Nageswara R Sastry <rnsastry@linux.ibm.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
Tested-by: Nageswara R Sastry <rnsastry@linux.ibm.com>
2023-03-06 07:39:27 -05:00
Mimi Zohar
2ea31a943c Update README
Update the README to reflect the changes to "evmctl --help".

Update the "--pass" option format in both the README and evmctl usage
to reflect passing an optional password on the command line (not
recommended).  When providing the password, the format is:
 "[--pass[=<password>]]".

Also fix some typos.

Still include references to both the deprecated "--rsa" and "--engine"
options.

Related confiigure options:
--enable-sigv1          Build ima-evm-utils with signature v1 support
--disable-engine        build ima-evm-utils without OpenSSL engine support

Reported-by: Vitaly Chikunov <vt@altlinux.org> # typos, "--pass" format
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2023-03-03 15:52:58 -05:00
Mimi Zohar
02c833339b Merge branch 'mmap-check-test' into next
- New ima_policy_check and mmap_check tests
- Ability to indicate missing kernel patch on test failure
2023-03-02 17:11:17 -05:00
Roberto Sassu
6917e384d3 Add tests for MMAP_CHECK and MMAP_CHECK_REQPROT hooks
Add tests to ensure that, after applying the kernel patch 'ima: Align
ima_file_mmap() parameters with mmap_file LSM hook', the MMAP_CHECK hook
checks the protections applied by the kernel and not those requested by the
application.

Also ensure that after applying 'ima: Introduce MMAP_CHECK_REQPROT hook',
the MMAP_CHECK_REQPROT hook checks the protections requested by the
application.

Test both with the test_mmap application that by default requests the
PROT_READ protection flag. Its syntax is:

test_mmap <file> <mode>

where mode can be:
- exec: adds the PROT_EXEC protection flag to mmap()
- read_implies_exec: calls the personality() system call with
                     READ_IMPLIES_EXEC as the first argument before mmap()
- mprotect: adds the PROT_EXEC protection flag to a memory area in addition
            to PROT_READ
- exec_on_writable: calls mmap() with PROT_EXEC on a file which has a
                    writable mapping

Check the different combinations of hooks/modes and ensure that a
measurement entry is found in the IMA measurement list only when it is
expected. No measurement entry should be found when only the PROT_READ
protection flag is requested or the matching policy rule has the
MMAP_CHECK_REQPROT hook and the personality() system call was called with
READ_IMPLIES_EXEC.

mprotect() with PROT_EXEC on an existing memory area protected with
PROT_READ should be denied (with an appraisal rule), regardless of the MMAP
hook specified in the policy. The same applies for mmap() with PROT_EXEC on
a file with a writable mapping.

Signed-off-by: Roberto Sassu <roberto.sassu@huawei.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2023-03-02 16:52:50 -05:00
Roberto Sassu
6a658e23d6 Add ima_policy_check.awk and ima_policy_check.test
Add ima_policy_check.awk to check for possible overlapping of a rule being
added by a test with the existing IMA policy (policy replacement by IMA at
the first policy load is not taken into account).

ima_policy_check.awk expects as input the rule to be added, followed by the
IMA policy.

It returns a bit mask with the following values:
- 1: invalid new rule;
- 2: overlap of the new rule with an existing rule in the IMA policy;
- 4: new rule exists in the IMA policy.

Values can be individually checked by the test executing the awk script, to
determine what to do (abort loading, print a warning in case of overlap,
avoid adding an existing rule).

The bit mask allows the test to see multiple statements regarding the new
rule. For example, if the test added anyway an overlapping rule, it could
also see that the policy already contains it at the next test execution,
and does not add it again.

Since ima_policy_check.awk uses GNU extensions (such as the or() function,
or the fourth argument of split()), add gawk as dependency for the CI.

Finally add ima_policy_check.test, to ensure that the awk script behaves as
expected.

Signed-off-by: Roberto Sassu <roberto.sassu@huawei.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2023-03-02 16:52:25 -05:00
Roberto Sassu
1d3a0b6923 Introduce expect_pass_if() and expect_fail_if()
Introduce these functions to let the developer specify which kernel patches
are required for the tests to be successful (either pass or fail). If a
test is not successful, print those patches in the test result summary.

First, the developer should declare an array, named PATCHES, with the list
of all kernel patches that are required by the tests. For example:

PATCHES=(
'patch 1 title'
...
'patch N title'
)

Second, the developer could replace the existing expect_pass() and
expect_fail() respectively with expect_pass_if() and expect_fail_if(), and
add the indexes in the PATCHES array as the first argument, enclosed with
quotes. The other parameters of expect_pass() and expect_fail() remain the
same.

In the following example, the PATCHES array has been added to a new test
script, tests/mmap_check.test:

PATCHES=(
'ima: Align ima_file_mmap() parameters with mmap_file LSM hook'
'ima: Introduce MMAP_CHECK_REQPROT hook'
)

Then, expect_pass() has been replaced with expect_pass_if():

expect_pass_if '0' check_mmap "MMAP_CHECK" "read_implies_exec"

The resulting output when a test fails (if the required patch is not
applied) is:

Test: check_mmap (hook="MMAP_CHECK", test_mmap arg: "read_implies_exec")
Result (expect found): not found
Possibly missing patches:
 - ima: Align ima_file_mmap() parameters with mmap_file LSM hook

Signed-off-by: Roberto Sassu <roberto.sassu@huawei.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2023-03-01 11:00:22 -05:00
Mimi Zohar
8f6ba073a0 Fix reading the TPM 2.0 PCRs
Prior to the support for reading the TPM 2.0 PCRs via the sysfs
interface, based on environment variables the userspace application read
either the physical or software TPM's PCRs.

With the support for reading the exported TPM 2.0 PCRs via the sysfs
interface, the physical TPM's PCRs are always read.  Define a new evmctl
option named '--hwtpm' to limit reading the TPM 2.0 PCRs via the sysfs
interface.

Fixes: a141bd594263 ("add support for reading per bank TPM 2.0 PCRs via sysfs")
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2023-02-23 09:37:54 -05:00
Eric Biggers
0290acff79 tests: use new git repo URL for fsverity-utils
See the announcement at "[ANNOUNCE] Moving the fsverity-utils git repo"
(https://lore.kernel.org/r/Y9GKm+hcm70myZkr@sol.localdomain).

Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2023-02-20 08:48:09 -05:00
Petr Vorel
d50e8c4397 github: Put openSSL build into own section
That helps readability when reviewing logs.

Signed-off-by: Petr Vorel <pvorel@suse.cz>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2023-02-15 16:48:52 -05:00
Petr Vorel
80442de4dd github: travis: Remove COMPILE_SSL from tumbleweed
Distro has openSSL 3.0.7, no need to compile own openSSL 3.x.

Signed-off-by: Petr Vorel <pvorel@suse.cz>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2023-02-15 16:48:52 -05:00
Petr Vorel
fdc2788d8f tests/install-swtpm.sh: Update ibmswtpm2 to 1682
At least on Tumbleweed build fails due openSSL 3.0.7
being installed from package.

Signed-off-by: Petr Vorel <pvorel@suse.cz>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2023-02-15 16:48:52 -05:00
Mimi Zohar
d18d6fff5c ci: cleanup build.sh test log output
Unlike the original ima-evm-utils ima_hash.test and sign_verify.test
selftests, kernel tests may fail for any number of reasons (e.g. kernel
config, permissions, missing applications, test infrastructure).  For
these tests, the full test log is needed to analyze the failure.

Create a phony target in tests/Makefile.am named "check-logs". Based on
test name, output different amounts of the test log.

Reviewed-by: Petr Vorel <pvorel@suse.cz>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2023-01-27 11:55:53 -05:00
Mimi Zohar
58b4c7ac4b Merge branch 'uml' into next
From "Support testing in new enviroments" cover letter:

One of the main limitations of running tests in the current environment is
that features/bug fixes to be tested need to be already included in the
running kernel, which is not always the case.

User Mode Linux (UML) and virtual machines can be used to overcome this
limitation. They allow to create a new environment and run a custom kernel
built by a CI or by the user. The tests can then check the features/bug
fixes of the custom kernel.

Running tests in a new environment also gives the ability to control the
configuration, and to have a clean state for each test by creating new
environments as necessary. The current environment might not allow that,
e.g. for security reasons.

Introduce a mechanism for creating and managing new environments. Expose an
API that allow to transparently create one or multiple environments in a
test script, and to reexecute that script in the new one. Using that API
requires minimal changes to the existing scripts.

The API is generic enough to support different types of enviroments. The
environment can be selected with the TST_ENV environment variable. At the
moment, only UML is supported. QEMU will be added at a later stage.

With the ability to test custom kernels, ima-evm-utils might introduce
specific tests for that, separated from the tests to verify the
ima-evm-utils user space functionality. At the moment, there is no such
distinction, existing tests verify both.

First, fix error messages and a variable in evmctl. Then, add kernel
configuration options for the tests, to be merged with the default
configuration. Add a new job in the Github workflow to build the UML kernel
from a repository and branch specified in the LINUX_URL and LINUX_BRANCH
variables (if the kernel repository does not have a branch with the same
name of the ima-evm-utils one). Per Github documentation, these variables
can be defined at organization, repository and environment level.

Return the correct script exit code if no test was executed. Introduce the
new API for creating and managing new enviroments, for existing and new
test scripts. If TST_ENV is not set, calling the API results in a nop, and
tests are executed in the current environment.

Add the possibility to select individual tests to run in a test script,
with the TST_LIST variable, so that a new environment can be created
multiple times for a subset of tests (useful if for example a test require
kernel settings different from the previous test).

Add tests for EVM portable signatures and modify fsverity.test to use the
new API.

Finally, don't require making changes to the system to run fsverity.test,
install a software dependency after the appropriate repository has been
set up, and temporarily remove CONFIG_DEBUG_SG to avoid a kernel panic
until the patches to fix it are accepted in the upstream kernel.
2023-01-27 11:49:19 -05:00
Mimi Zohar
40962a6690 Temporarily remove CONFIG_DEBUG_SG to test portable signatures
Enabling CONFIG_DEBUG_SG requires two kernel fixes. For now don't
enable CONFIG_DEBUG_SG.

Fixes: a910fe25a975 ("Add kernel configuration for tests")
Signed-off-by: Roberto Sassu <roberto.sassu@huawei.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2023-01-27 11:43:41 -05:00
Mimi Zohar
f3289d5598 ci: haveged requires EPEL on CentOS stream:8
The travis "fedora:latest" matrix rule fails due to not finding
"haveged".  Install "haveged" after enabling EPEL.

Fixes: f106a9022d1f ("Add support for creating a new testing environment in functions.sh")
Reviewed-by: Stefan Berger <stefanb@linux.ibm.com>
Signed-off-by: Roberto Sassu <roberto.sassu@huawei.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2023-01-27 11:42:44 -05:00
Roberto Sassu
452f4b2eac Use in-place built fsverity binary instead of installing it
Instead of making changes to the system, use in-place built fsverity binary
by adding ../fsverity-utils to the PATH variable, so that the binary can be
found with the 'command -v' command.

Don't delete the fsverity-utils directory, so that the built binary is
available. Not deleting should not be a problem, as the script is meant to
be executed in a CI environment, where cleanup is done by the CI
infrastructure itself.

Signed-off-by: Roberto Sassu <roberto.sassu@huawei.com>
Reviewed-by: Stefan Berger <stefanb@linux.ibm.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2023-01-27 11:40:31 -05:00
Roberto Sassu
0bccb5412c Adapt fsverity.test to be able to run in a new testing environment
Adapt fsverity.test by adding calls to the testing environment API in
functions.sh. If TST_ENV is set, create a new environment and run the
kernel specified with the TST_KERNEL environment variable. Otherwise, keep
the current behavior.

Signed-off-by: Roberto Sassu <roberto.sassu@huawei.com>
Reviewed-by: Stefan Berger <stefanb@linux.ibm.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2023-01-27 11:40:31 -05:00
Roberto Sassu
cf832d72f9 Add tests for EVM portable signatures
Verify that operations on files with EVM portable signatures succeed and
that the new kernel patch set does not break the existing kernel integrity
expectations. Build and install mount-idmapped for ci/fedora.sh, to
additionally test idmapped mounts.

To run the tests, pass the path of the kernel private key with the
TST_KEY_PATH environment variable. If not provided, search first in the
ima-evm-utils top directory, and then in
/lib/modules/$(uname -r)/source/certs/signing_key.pem and
/lib/modules/$(uname -r)/build/certs/signing_key.pem.

Root privileges are required to mount the image, configure IMA/EVM and set
xattrs.

Set TST_ENV to 'um', to relaunch the script in a new environment after
booting an UML kernel. The UML kernel path must be specified with the
TST_KERNEL environment variable.

Alternatively, set the TST_EVM_CHANGE_MODE variable to 1, to change the
current EVM mode, if a test needs a different one. Otherwise, execute only
the tests compatible with the current EVM mode.

Also set the EVM_ALLOW_METADATA_WRITES flag in the EVM mode, before
launching the script, to run the check_evm_revalidate() test. Execute:

echo 4 > /sys/kernel/security/evm

The last two environment variables above affect which tests will run the
next time the script is executed. Without setting TST_ENV, changes to the
current EVM mode will be irreversibly done in the host. Next time, unless
the host is rebooted, only tests compatible with the last EVM mode set will
run. The others will be skipped.

By setting TST_ENV, this problem does not arise as, every time the
environment is created, it will be clean with no flags set in the EVM mode.

Signed-off-by: Roberto Sassu <roberto.sassu@huawei.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2023-01-27 11:40:31 -05:00
Roberto Sassu
b573b7d4a1 Introduce TST_LIST variable to select a test to execute
It might be desirable, due to restrictions in the testing environment, to
execute tests individually. Introduce the TST_LIST variable, which can be
set with the name of the test to execute. If the variable is set,
expect_pass and expect_fail automatically skip the tests when the first
argument of those functions does not match the value of TST_LIST.

TST_LIST can be also used in new environments, to execute a subset of
defined tests for each environment. It is sufficient to add the variable
and its value to the kernel command line.

Signed-off-by: Roberto Sassu <roberto.sassu@huawei.com>
Reviewed-by: Stefan Berger <stefanb@linux.ibm.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2023-01-27 11:40:31 -05:00
Roberto Sassu
f106a9022d Add support for creating a new testing environment in functions.sh
Add the new functions _run_env(), _exit_env(), _init_env() and
_cleanup_env() to run the tests inside a new environment specified with the
TST_ENV environment variable.

A typical structure of a script with tests is:

trap '_report_exit_and_cleanup _cleanup_env cleanup' \
    SIGINT SIGTERM SIGSEGV EXIT

cleanup() {
	<test cleanup>
}

<tests implementations>

_run_env "$TST_KERNEL" "$PWD/$(basename "$0")" "env_var1=$env_var1 ..."

_exit_env "$TST_KERNEL"

_init_env

<tests init>

<tests call>

If TST_ENV is not set or empty, don't create a new testing environment and
perform the cleanup in the current environment. Don't create a new testing
environment also if the script is already executed in a new environment, to
avoid loops. Instead, for cleanup, do it in the new environment and skip it
in the host environment (if the cleanup function is passed to
_cleanup_env()).

Signal to the creator of the environment failures of tests or of the script
itself run in the new environment (if the exit code is 1 ($FAIL) or 99
($HARDFAIL)) with an unclean shutdown of the system.

Add haveged and systemd as dependencies for the tests in ci/fedora.sh,
respectively for initializing the random number generator and for shutting
down the system in the new environment.

Signed-off-by: Roberto Sassu <roberto.sassu@huawei.com>
Reviewed-by: Stefan Berger <stefanb@linux.ibm.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2023-01-27 11:40:16 -05:00
Roberto Sassu
03b5d159ca Pass cleanup function and its arguments to _report_exit_and_cleanup()
If an error occurs before any test is executed, _report_exit_and_cleanup()
returns 77 ($SKIP) as exit code, which might not reflect the real exit code
at the time the script terminated its execution.

If the function registered in the shell trap() is a cleanup function
calling _report_exit_and_cleanup() inside, the latter will not have access
to the exit code at the time of the trap but instead to the exit code of
the cleanup function.

To solve this issue, pass the cleanup function and its arguments to
_report_exit_and_cleanup(), so that the latter can first get the script
exit code and then can execute the cleanup function.

Finally, if no test was executed, return the exit code at the time of the
trap() instead of 77.

Signed-off-by: Roberto Sassu <roberto.sassu@huawei.com>
Reviewed-by: Stefan Berger <stefanb@linux.ibm.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2023-01-27 11:39:24 -05:00
Roberto Sassu
3fadf997a6 Compile the UML kernel and download it in Github Actions
Add a build job, prerequisite of the existing job, to compile the UML
kernel and upload it and the signing key to a cache. Github configuration
should have two variables: LINUX_URL, the full URL of the kernel
repository; LINUX_BRANCH, the branch to check out as fallback if the kernel
repository does not have the same branch name as the one being pushed for
ima-evm-utils. See:

https://docs.github.com/en/actions/learn-github-actions/variables

for directions on how to define those variables.

If the two variables are not defined, the default values are:

LINUX_URL=https://git.kernel.org/pub/scm/linux/kernel/git/zohar/linux-integrity.git
LINUX_BRANCH=next-integrity

If there is a cache hit (same kernel commit and same kernel configuration),
next time the UML kernel will not be rebuilt. To use the cache, it is
necessary to install zstd in the container. Add this dependency to
ci/fedora.sh.

The cache can be managed at the following URL:

https://github.com/<username>/ima-evm-utils/actions/caches

The page also offers the possibility to clean the cache, to force
rebuilding the kernel.

Add a new entry in the testing matrix, for the fedora-latest container
image, to run the tests with the UML kernel. The entry differs from the
others for the new environment variable TST_ENV, set to 'um', and
TST_KERNEL set to '../linux', as the tests will be executed from the
tests/ directory in ima-evm-utils.

Add a new volume to the container, /dev/shm from the host, as it is
required for running the UML kernel.

Extend the existing job with steps to download the UML kernel and signing
key from the cache. The new steps are executed only if the matrix entry has
TST_ENV set.

Finally, pass TST_ENV and TST_KERNEL to the tests. A test should also
propagate these variables to the new environment, by passing them to the
kernel command line.

Signed-off-by: Roberto Sassu <roberto.sassu@huawei.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2023-01-27 11:39:24 -05:00
Roberto Sassu
a910fe25a9 Add kernel configuration for tests
Add kernel-configs/base with changes to be applied to the default kernel
configuration, generated with 'make defconfig'.

Add kernel-configs/integrity, with integrity-specific configuration
options.

Splitting changes helps to identify more easily the desired group of
options. In the future, options could be split even further.

All changes in this directory will be applied with the merge_config.sh
script from the kernel source code in a Github workflow step.

Signed-off-by: Roberto Sassu <roberto.sassu@huawei.com>
Reviewed-by: Stefan Berger <stefanb@linux.ibm.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2023-01-27 11:38:33 -05:00
Roberto Sassu
d1b48e9783 Fix error messages and vars in calc_evm_hmac()
Make sure that the function name in the error message corresponds to the
actual function called.

Rename mdlen and hash respectively to siglen and sig. Also, initialize
siglen to the size of sig (MAX_DIGEST_SIZE), as this is recommended in the
documentation of EVP_DigestSignFinal().

Signed-off-by: Roberto Sassu <roberto.sassu@huawei.com>
Reviewed-by: Stefan Berger <stefanb@linux.ibm.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2023-01-27 11:37:20 -05:00
Alberto Mardegan
eea9827d99 libimaevm: do not crash if the certificate cannot be read
This code path can be triggered if someone inadvertedly swaps the key
with the certificate in the evmctl command line. Our `x` variable would
be NULL, and we need to abort further processing of the certificate.

Signed-off-by: Alberto Mardegan <a.mardegan@omp.ru>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2023-01-27 11:28:34 -05:00
Vitaly Chikunov
3f162e8e3d Experimental fsverity.test related GA CI improvements
This does not make fsverity.test working on GA CI, though.

- `--device /dev/loop-control' is required for losetup(8) to work.
- `--privileged' is required foo mount(8) to work, and this makes
  `--security-opt seccomp=unconfined' redundant.
- GA container does not have `/sys/kernel/security' mounted which is
  needed for `/sys/kernel/security/integrity/ima/policy'.
- Enable `set -x` in CI as the logs is everything we have to analyze on
  failures.

Update: with these changes and the UML kernel support, the fsverity.test
is working properly.

Signed-off-by: Vitaly Chikunov <vt@altlinux.org>
[zohar@linux.ibm.com: updated patch description]
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2023-01-27 11:28:32 -05:00
Mimi Zohar
b259a2ba8b tests: add fsverity measurement test
Test IMA support for including fs-verity enabled file measurements
in the IMA measurement list based on the ima-ngv2 and ima-sigv2
records.

Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2023-01-27 11:28:29 -05:00
Mimi Zohar
aad5d334a6 Save ima-evm-utils sourceforge wiki
The sourceforge wiki info is dated and requires a major overhaul.  Some
of the information already exists in the linux kernel documentation.
For now, save it with the referenced html files.

Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2023-01-27 11:28:07 -05:00
Stefan Berger
066685d162 Change condition to free(pub)
Change the condition under which pub is freed to make it clearer for the
reader and analyzer.

This change gets rid of the following gcc -fanalyzer warning:

evmctl.c:1140:12: warning: leak of ‘pub’ [CWE-401] [-Wanalyzer-malloc-leak]
 1140 |         if (imaevm_params.x509)
      |            ^

Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2022-11-15 18:18:19 -05:00
Stefan Berger
c7928795cd Add assert to ensure that algo_name in bank is set
To avoid numerous warning messages from gcc 12.2.1 when compiling with
-fanalyzer, insert an assert to ensure that algo_name in each bank
is set. The assert resolves the following warnings:

evmctl.c:1998:30: warning: use of NULL where non-null expected [CWE-476] [-Wanalyzer-null-argument]
 1998 |                         if (!strcmp(tpm_banks[j].algo_name, alg)) {
      |                              ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

evmctl.c: In function ‘ima_measurement’:
evmctl.c:2146:24: warning: use of NULL where non-null expected [CWE-476] [-Wanalyzer-null-argument]
 2146 |                     && strcmp(pseudo_padded_banks[c].algo_name, verify_bank)) {
      |                        ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  ‘ima_measurement’: events 1-2

evmctl.c: In function ‘cmd_ima_bootaggr’:
evmctl.c:2611:33: warning: use of NULL where non-null expected [CWE-476] [-Wanalyzer-null-argument]
 2611 |                 bootaggr_len += strlen(tpm_banks[i].algo_name) + 1;
      |                                 ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2022-11-15 18:18:19 -05:00
Stefan Berger
ca68ddd857 Fix memory leak related to entry.template
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2022-11-15 18:18:19 -05:00
Stefan Berger
d7dffec5f7 Fix memory leaks of tpm_bank_info allocations
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2022-11-15 18:18:19 -05:00
Tergel Myanganbayar
a141bd5942 add support for reading per bank TPM 2.0 PCRs via sysfs
Until Linux kernel version 5.11, a TSS was required to read TPM 2.0 PCR
values. A feature which exposed the per bank TPM 2.0 PCRs directly via
sysfs was upstreamed in newer Kernel versions.

Use this recent feature in IMA-EVM-UTILS to remove TSS dependency.

Signed-off-by: Tergel Myanganbayar <tergel@linux.ibm.com>
Reviewed-by: Stefan Berger <stefanb@linux.ibm.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2022-11-13 16:56:18 -05:00
Mimi Zohar
7aaf923d0b Fix tpm2_pcr_supported() output messages
Remove unnecessary path check in pcr_ibmtss.c and update the syntax
in the other.

Reviewed-by: Stefan Berger <stefanb@linux.ibm.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2022-11-13 16:56:18 -05:00
Mimi Zohar
27e91006d8 Sanity check the template data field sizes
The field sizes of the original "ima" template data are static, but
the other template data fields are not.  They're prefixed with a size.

Add some data field size sanity checks in ima_show_ng().

Reviewed-by: Stefan Berger <stefanb@linux.ibm.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2022-11-13 16:56:18 -05:00
Mimi Zohar
22f8effda5 Define and verify the template data length upper bounds
The template data length is variable, based on the template format.
Define some sort of upper bounds.

Reviewed-by: Stefan Berger <stefanb@linux.ibm.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2022-11-13 16:56:18 -05:00
Mimi Zohar
6778e3511b Don't ignore number of items read
fread() either returns the number of bytes read or the number of items
of data read.  Check that it returns the requested number of items read.

Reviewed-by: Stefan Berger <stefanb@linux.ibm.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2022-11-13 16:56:18 -05:00
Mimi Zohar
c8b1757270 Make sure the key file is a regular file
Before attempting to use the key file, make sure it is a regular file.

Reviewed-by: Stefan Berger <stefanb@linux.ibm.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2022-11-13 16:56:18 -05:00
Mimi Zohar
297d01bdb6 Build OpenSSL without engine support
Fix COMPILE_SSL to build for the proper architecture, link with the
appropriate library, and set up library path for evmctl.

Compile OpenSSL with "no-engine" and "no-dynamic-engine" support.

Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2022-11-13 16:56:18 -05:00
Mimi Zohar
abf7b5e236 Compile a newer version of OpenSSL
With the distros shipping OpenSSL 3 with engine support, the original
purpose for compiling OpenSSL 3 to test sm2/sm3 is no longer necessary
and could be removed.  Or, it could be re-purposed for building OpenSSL
without engine support, which is needed for testing.

For both travis and github actions, update openssl-3.0.0-beta1 with
openssl-3.0.5.

Reviewed-by: Stefan Berger <stefanb@linux.ibm.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2022-11-13 16:56:18 -05:00
Mimi Zohar
1d4970b46b Base sm2/sm3 test on openssl version installed
Since the distros are now shipping with OpenSSL 3, no need
to build it.  Limit the sm2/sm3 test to OpenSSL 3.

Reviewed-by: Stefan Berger <stefanb@linux.ibm.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2022-11-13 16:56:18 -05:00
Mimi Zohar
f57ea92d6e Missing template data size lower bounds checking
Each record in the IMA measurement list must contain some template data.
Ensure the template data is not zero length.

Reviewed-by: Petr Vorel <pvorel@suse.cz>
Reviewed-by: Stefan Berger <stefanb@linux.ibm.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2022-11-13 16:56:18 -05:00
Mimi Zohar
232836a079 Limit the file hash algorithm name length
Instead of assuming the file hash algorithm is a properly NULL terminated
string, properly limit the "algo:<hash>" field size.

Reviewed-by: Petr Vorel <pvorel@suse.cz>
Reviewed-by: Stefan Berger <stefanb@linux.ibm.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2022-11-13 16:56:18 -05:00
Mimi Zohar
f2b1b66b7c Fix potential use after free in read_tpm_banks()
On failure to read TPM 2.0 bank PCRs 'errmsg' is not properly set to
NULL after being freed.  Fix potential use after free.

Fixes: 3472f9ba9c05 ("ima-evm-utils: read the PCRs for the requested TPM banks")
Reviewed-by: Petr Vorel <pvorel@suse.cz>
Reviewed-by: Stefan Berger <stefanb@linux.ibm.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2022-11-13 16:56:18 -05:00
Mimi Zohar
c1635add22 Disable use of OpenSSL "engine" support
OpenSSL v3 "engine" support is deprecated and replaced with "providers".
Engine support will continue to work for a while, but results in
deprecated declaration and other messages.  One option is simply to hide
them ("-Wno-deprecated-declarations").  The other alternative is to
conditionally build ima-evm-utils without OpenSSL engine support and
without disabling deprecated declarations.

Based on "--disable-engine" or "--enable-engine=no" configuration
option, disable OpenSSL "engine" support.

As suggested by Vitaly,
- verify ENGINE_init symbol is defined in libcrypto
- disable engine support if either OPENSSL_NO_DYNAMIC_ENGINE or
OPENSSL_NO_ENGINE variables are defined

Reviewed-by: Stefan Berger <stefanb@linux.ibm.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2022-11-13 16:56:18 -05:00
Mimi Zohar
a7b5bdbf36 Add missing EVP_MD_CTX_free() call in calc_evm_hash()
When EVP_MD_CTX_new() call was added, the corresponding EVP_MD_CTX_free()
was never called.  Properly free it.

Fixes: 81010f0d87ef ("ima-evm-utils: Add backward compatible support for openssl 1.1")
Reviewed-by: Petr Vorel <pvorel@suse.cz>
Reviewed-by: Stefan Berger <stefanb@linux.ibm.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2022-11-13 16:56:18 -05:00
Mimi Zohar
67ca790435 Replace the low level HMAC calls when calculating the EVM HMAC
Calculating the EVM HMAC and labeling the filesystem was originally
included in ima-evm-utils for debugging purposes only.  For now,
instead of removing EVM HMAC support just replace the low level
HMAC_ calls with EVP_ calls.

The '-a, --hashalgo' specifies the IMA hash or signature algorithm.
The kernel EVM HMAC is limited to SHA1.  Fix ima-evm-utils by hard
coding the EVM HMAC algorithm to SHA1.

Reviewed-by: Petr Vorel <pvorel@suse.cz>
Reviewed-by: Stefan Berger <stefanb@linux.ibm.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2022-11-13 16:56:18 -05:00
Mimi Zohar
b9c9759a7e Replace the low level SHA1 calls when calculating the TPM 1.2 PCRs
OpenSSL v3 emits deprecated warnings for SHA1 functions.  Use the
EVP_ functions when walking the TPM 1.2 binary bios measurements
to calculate the TPM 1.2 PCRs.

Reviewed-by: Petr Vorel <pvorel@suse.cz>
Reviewed-by: Stefan Berger <stefanb@linux.ibm.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2022-11-13 16:56:18 -05:00
Mimi Zohar
751a395772 Deprecate IMA signature version 1
The original IMA file signatures were based on a SHA1 hash.  Kernel
support for other hash algorithms was subsequently upstreamed.  Deprecate
"--rsa" support.

Define "--enable-sigv1" option to configure signature v1 support.

Reviewed-by: Petr Vorel <pvorel@suse.cz>
Reviewed-by: Stefan Berger <stefanb@linux.ibm.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2022-11-13 16:56:18 -05:00
Mimi Zohar
8e1da33b0c Update configure.ac to address a couple of obsolete warnings
Remove AC_PROG_LIBTOOL and AC_HEAD_STDC. Replace AC_HELP_STRING with
AS_HELP_STRING.

Reviewed-by: Petr Vorel <pvorel@suse.cz>
Reviewed-by: Stefan Berger <stefanb@linux.ibm.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2022-11-13 16:56:18 -05:00
Mimi Zohar
f8c9621d88 travis: update dist=focal
Although Github Actions is available on Github Enterprise Server 3.x
single server edition, as well as the unpaid version, it is not
available in Github Enterprise Server 3.x cluster edition[1].

Continue updating travis.yml.

[1] https://docs.github.com/en/enterprise-server@3.0/admin/release-notes#github-packages

Reviewed-by: Petr Vorel <pvorel@suse.cz>
Reviewed-by: Stefan Berger <stefanb@linux.ibm.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2022-11-13 16:56:18 -05:00
Mimi Zohar
1fcac50e30 Log and reset 'errno' on lsetxattr failure
Writing either security.ima hashes or security.evm hmacs from userspace
will fail regardless of the IMA or EVM fix mode.  In fix mode, 'touch'
will force security.ima and security.evm to be updated.

Make the setxattr error messages more explicit and clear errno.

Reviewed-by: Stefan Berger <stefanb@linux.ibm.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2022-11-13 16:56:11 -05:00
Mimi Zohar
ba2b6a97c1 log and reset 'errno' after failure to open non-critical files
Define a log_errno_reset macro to emit the errno string at or near the
time of error, similar to the existing log_errno macro, but also reset
errno to avoid dangling or duplicate errno messages on exit.

The initial usage is for non-critical file open failures.

Suggested-by: Vitaly Chikunov <vt@altlinux.org>
Reviewed-by: Vitaly Chikunov <vt@altlinux.org>
Reviewed-by: Petr Vorel <pvorel@suse.cz>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2022-11-13 16:55:50 -05:00
Mimi Zohar
0f3b9a0b2c Revert "Reset 'errno' after failure to open or access a file"
This reverts commit acb19d1894a4a95471b8d2346cd6c3ecf3385110, based on
the mailing list discussion and will be fixed in the next commit.

Reviewed-by: Stefan Berger <stefanb@linux.ibm.com>
Reviewed-by: Petr Vorel <pvorel@suse.cz>
Link: https://lore.kernel.org/linux-integrity/20220915153659.dtykhzitxdrlpasq@altlinux.org/
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2022-11-06 18:31:31 -05:00
Petr Vorel
75fadad261 ci/alpine.sh: Install bash
bash is a dependency for tests, not being installed by default on
containers.

This fixes:
../test-driver: line 112: ./ima_hash.test: not found
../test-driver: line 112: ./sign_verify.test: not found
../test-driver: line 112: ./boot_aggregate.test: not found

Signed-off-by: Petr Vorel <pvorel@suse.cz>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2022-08-29 17:55:57 -04:00
Petr Vorel
8f1e5224e6 ci/ubuntu: impish -> jammy
Ubuntu 21.10 impish EOL in 2022-04 (next month).
Replace it with the latest stable release (EOL 2027-04).

Signed-off-by: Petr Vorel <pvorel@suse.cz>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2022-08-29 17:55:57 -04:00
Mimi Zohar
3d77138811 Verify an fs-verity file digest based signature
ima-evm-utils does not attempt to calculate or even read the fs-verity
file hash, but can verify the fs-verity signature based on the fsverity
file hash, both contained in the measurement list record.

Example:
evmctl ima_measurement --key <DER encoded public key> \
 --verify-sig /sys/kernel/security/ima/binary_runtime_measurements

Modify 'sig' argument of verify_hash() to be the full xattr in order to
differentiate signatures types.

Note:
Kernel commit b1aaab22e263 ("ima: pass full xattr with the signature")
added the 'type' to signature_v2_hdr struct, which hasn't been reflected
here. (todo)

Reviewed-by: Stefan Berger <stefanb@linux.ibm.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2022-08-29 17:55:57 -04:00
Mimi Zohar
fc46af121e Sign an fs-verity file digest
Sign fs-verity file digests provided in the format as produced by
"fsverity digest".  The output is of the same format as the input,
but with the file signature appended.  Use setfattr to write the
signature as security.ima xattr.

fsverity digest format: <algo>:<hash> <pathname>
output format: <algo>:<hash> <pathname> <signature>

Instead of directly signing the fsverity hash, to disambiguate the
original IMA signatures from the fs-verity signatures stored in the
security.ima xattr a new signature format version 3 (sigv3) was
defined as the hash of the xattr type (enum evm_ima_xattr_type),
the hash algorithm (enum hash_algo), and the hash.

Example:
fsverity digest <pathname> | evmctl sign_hash --veritysig \
 --key <pem encoded private key>

Reviewed-by: Stefan Berger <stefanb@linux.ibm.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2022-08-29 17:55:46 -04:00
Mimi Zohar
acb19d1894 Reset 'errno' after failure to open or access a file
Not being able to open a file is not necessarily a problem. If
and when it occurs, an informational or error message with the
actual filename is emitted as needed.

Reset 'errno' to prevent the "errno: No such file or directory (2)"
generic message.

Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2022-08-09 10:34:25 -04:00
Mimi Zohar
eb956b8d35 travis: install fuse-overlayfs before podman
WARN[0000] Error validating CNI config file /home/travis/.config/cni/net.d/87-podman.conflist: [failed to find plugin "bridge" in path [/usr/local/libexec/cni /usr/libexec/cni /usr/local/lib/cni /usr/lib/cni /opt/cni/bin] failed to find plugin "portmap" in path [/usr/local/libexec/cni /usr/libexec/cni /usr/local/lib/cni /usr/lib/cni /opt/cni/bin] failed to find plugin "firewall" in path [/usr/local/libexec/cni /usr/libexec/cni /usr/local/lib/cni /usr/lib/cni /opt/cni/bin] failed to find plugin "tuning" in path [/usr/local/libexec/cni /usr/libexec/cni /usr/local/lib/cni /usr/lib/cni /opt/cni/bin]]

Based on https://github.com/containers/podman/blob/main/docs/tutorials/rootless_tutorial.md, install
fuse-overlayfs.

Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2022-06-14 07:53:27 -04:00
Mimi Zohar
170be44a7b travis: include CentOS stream 8
Replace CentOS 8 with CentOS stream 8.
Use podman for both CentOS 7 & 8.

Reviewed-by: Petr Vorel <pvorel@suse.cz>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2022-03-24 10:34:23 -04:00
Petr Vorel
e06980b245 ci/GitHub: Remove CentOS 8
It EOL in 12/2021 and CI is failing due removed repo:

CentOS Linux 8 - AppStream                      232  B/s |  38  B     00:00
Error: Failed to download metadata for repo 'appstream': Cannot prepare internal mirrorlist: No URLs in mirrorlist

Removing only from GitHub Actions, because Mimi Zohar reported Travis
can use centos:stream8.

Signed-off-by: Petr Vorel <pvorel@suse.cz>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2022-03-23 14:26:51 -04:00
Petr Vorel
37317838b4 ci: Replace groovy -> impish
in both GitHub Actions and Travis.

groovy is EOL, which is probably the reason why it's archives has been
removed:

Ign:1 http://security.ubuntu.com/ubuntu groovy-security InRelease
Ign:2 http://archive.ubuntu.com/ubuntu groovy InRelease
Err:3 http://security.ubuntu.com/ubuntu groovy-security Release
  404  Not Found [IP: 91.189.91.39 80]
Ign:4 http://archive.ubuntu.com/ubuntu groovy-updates InRelease
Ign:5 http://archive.ubuntu.com/ubuntu groovy-backports InRelease
Err:6 http://archive.ubuntu.com/ubuntu groovy Release
  404  Not Found [IP: 91.189.88.142 80]
Err:7 http://archive.ubuntu.com/ubuntu groovy-updates Release
  404  Not Found [IP: 91.189.88.142 80]
Err:8 http://archive.ubuntu.com/ubuntu groovy-backports Release
  404  Not Found [IP: 91.189.88.142 80]
Reading package lists...
E: The repository 'http://security.ubuntu.com/ubuntu groovy-security Release' does not have a Release file.
E: The repository 'http://archive.ubuntu.com/ubuntu groovy Release' does not have a Release file.
E: The repository 'http://archive.ubuntu.com/ubuntu groovy-updates Release' does not have a Release file.
E: The repository 'http://archive.ubuntu.com/ubuntu groovy-backports Release' does not have a Release file.

Using impish requires to use workaround to avoid apt asking to
interactively configure tzdata.

Signed-off-by: Petr Vorel <petr.vorel@gmail.com>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2022-01-19 16:34:42 -05:00
43 changed files with 4987 additions and 220 deletions

View File

@ -3,7 +3,79 @@ name: "distros"
on: [push, pull_request]
jobs:
build:
runs-on: ubuntu-latest
outputs:
LINUX_SHA: ${{ steps.last-commit.outputs.LINUX_SHA }}
name: build
timeout-minutes: 100
strategy:
fail-fast: false
steps:
- uses: actions/checkout@v3
- name: Determine last kernel commit
id: last-commit
shell: bash
run: |
mkdir linux-integrity
pushd linux-integrity
git init
LINUX_URL=${{ vars.LINUX_URL }}
if [ -z "$LINUX_URL" ]; then
LINUX_URL=https://git.kernel.org/pub/scm/linux/kernel/git/zohar/linux-integrity.git
fi
LINUX_BRANCH=${{ vars.LINUX_BRANCH }}
if [ -z "$LINUX_BRANCH" ]; then
LINUX_BRANCH=next-integrity
fi
git remote add origin $LINUX_URL
LINUX_SHA=$(git ls-remote origin $GITHUB_REF_NAME | awk '{print $1}')
[ -z "$LINUX_SHA" ] && LINUX_SHA=$(git ls-remote origin $LINUX_BRANCH | awk '{print $1}')
echo "LINUX_SHA=$LINUX_SHA" >> $GITHUB_OUTPUT
popd
- name: Cache UML kernel
id: cache-linux
uses: actions/cache@v3
with:
path: linux
key: linux-${{ steps.last-commit.outputs.LINUX_SHA }}-${{ hashFiles('**/kernel-configs/*') }}
- name: Cache signing key
id: cache-key
uses: actions/cache@v3
with:
path: signing_key.pem
key: signing_key.pem-${{ steps.last-commit.outputs.LINUX_SHA }}-${{ hashFiles('**/kernel-configs/*') }}
- name: Compile UML kernel
if: steps.cache-linux.outputs.cache-hit != 'true' || steps.cache-key.outputs.cache-hit != 'true'
shell: bash
run: |
if [ "$DEVTOOLSET" = "yes" ]; then
source /opt/rh/devtoolset-10/enable
fi
if [ "$ARCH" = "i386" ]; then
CROSS_COMPILE_OPT="CROSS_COMPILE=i686-linux-gnu-"
fi
pushd linux-integrity
git pull --depth 1 origin ${{ steps.last-commit.outputs.LINUX_SHA }}
make ARCH=um defconfig
./scripts/kconfig/merge_config.sh -m .config $(ls ../kernel-configs/*)
# Update manually, to specify ARCH=um
make ARCH=um olddefconfig
# Make everything built-in
make ARCH=um localyesconfig
make ARCH=um $CROSS_COMPILE_OPT -j$(nproc)
chmod +x linux
cp linux ..
cp certs/signing_key.pem ..
popd
job:
needs: build
runs-on: ubuntu-latest
strategy:
@ -17,7 +89,7 @@ jobs:
ARCH: i386
TSS: tpm2-tss
VARIANT: i386
COMPILE_SSL: openssl-3.0.0-beta1
COMPILE_SSL: openssl-3.0.5
# cross compilation builds
- container: "debian:stable"
@ -52,18 +124,17 @@ jobs:
env:
CC: clang
TSS: ibmtss
COMPILE_SSL: openssl-3.0.0-beta1
- container: "opensuse/leap"
env:
CC: gcc
TSS: tpm2-tss
- container: "ubuntu:groovy"
- container: "ubuntu:jammy"
env:
CC: gcc
TSS: ibmtss
COMPILE_SSL: openssl-3.0.0-beta1
COMPILE_SSL: openssl-3.0.5
- container: "ubuntu:xenial"
env:
@ -75,12 +146,14 @@ jobs:
CC: clang
TSS: ibmtss
- container: "centos:7"
- container: "fedora:latest"
env:
CC: gcc
TSS: tpm2-tss
CC: clang
TSS: ibmtss
TST_ENV: um
TST_KERNEL: ../linux
- container: "centos:latest"
- container: "centos:7"
env:
CC: gcc
TSS: tpm2-tss
@ -103,7 +176,7 @@ jobs:
container:
image: ${{ matrix.container }}
env: ${{ matrix.env }}
options: --security-opt seccomp=unconfined
options: --privileged --device /dev/loop-control -v /dev/shm:/dev/shm
steps:
- name: Show OS
@ -119,7 +192,12 @@ jobs:
INSTALL="${INSTALL%%/*}"
if [ "$VARIANT" ]; then ARCH="$ARCH" ./ci/$INSTALL.$VARIANT.sh; fi
ARCH="$ARCH" CC="$CC" TSS="$TSS" ./ci/$INSTALL.sh
if [ "$COMPILE_SSL" ]; then COMPILE_SSL="$COMPILE_SSL" ./tests/install-openssl3.sh; fi
- name: Build openSSL
run: |
if [ "$COMPILE_SSL" ]; then
COMPILE_SSL="$COMPILE_SSL" VARIANT="$VARIANT" ./tests/install-openssl3.sh; \
fi
- name: Build swtpm
run: |
@ -130,8 +208,24 @@ jobs:
fi
fi
- name: Retrieve UML kernel
if: ${{ matrix.env.TST_ENV }}
uses: actions/cache@v3
continue-on-error: false
with:
path: linux
key: linux-${{ needs.build.outputs.LINUX_SHA }}-${{ hashFiles('**/kernel-configs/*') }}
- name: Retrieve signing key
if: ${{ matrix.env.TST_ENV }}
continue-on-error: false
uses: actions/cache@v3
with:
path: signing_key.pem
key: signing_key.pem-${{ needs.build.outputs.LINUX_SHA }}-${{ hashFiles('**/kernel-configs/*') }}
- name: Compiler version
run: $CC --version
- name: Compile
run: CC="$CC" VARIANT="$VARIANT" ./build.sh
run: CC="$CC" VARIANT="$VARIANT" COMPILE_SSL="$COMPILE_SSL" TST_ENV="$TST_ENV" TST_KERNEL="$TST_KERNEL" ./build.sh

View File

@ -1,6 +1,6 @@
# Copyright (c) 2017-2021 Petr Vorel <pvorel@suse.cz>
dist: bionic
dist: focal
language: C
services:
- docker
@ -9,7 +9,7 @@ matrix:
include:
# 32 bit build
- os: linux
env: DISTRO=debian:stable VARIANT=i386 ARCH=i386 TSS=tpm2-tss COMPILE_SSL=openssl-3.0.0-beta1
env: DISTRO=debian:stable VARIANT=i386 ARCH=i386 TSS=tpm2-tss COMPILE_SSL=openssl-3.0.5
compiler: gcc
# cross compilation builds
@ -32,7 +32,7 @@ matrix:
# glibc (gcc/clang)
- os: linux
env: DISTRO=opensuse/tumbleweed TSS=ibmtss CONTAINER=podman CONTAINER_ARGS="--runtime=/usr/bin/crun --network=host" COMPILE_SSL=openssl-3.0.0-beta1
env: DISTRO=opensuse/tumbleweed TSS=ibmtss CONTAINER=podman CONTAINER_ARGS="--runtime=/usr/bin/crun --network=host"
compiler: clang
- os: linux
@ -40,7 +40,7 @@ matrix:
compiler: gcc
- os: linux
env: DISTRO=ubuntu:groovy TSS=ibmtss COMPILE_SSL=openssl-3.0.0-beta1
env: DISTRO=ubuntu:jammy TSS=ibmtss COMPILE_SSL=openssl-3.0.5
compiler: gcc
- os: linux
@ -52,11 +52,11 @@ matrix:
compiler: clang
- os: linux
env: DISTRO=centos:7 TSS=tpm2-tss
env: DISTRO=centos:7 TSS=tpm2-tss CONTAINER=podman CONTAINER_ARGS="--runtime=/usr/bin/crun --network=host"
compiler: gcc
- os: linux
env: DISTRO=centos:latest TSS=tpm2-tss
env: REPO="quay.io/centos/" DISTRO="${REPO}centos:stream8" TSS=tpm2-tss CONTAINER=podman CONTAINER_ARGS="--runtime=/usr/bin/crun --network=host"
compiler: clang
- os: linux
@ -82,7 +82,7 @@ before_install:
sudo sh -c "echo 'deb http://download.opensuse.org/repositories/devel:/kubic:/libcontainers:/stable/xUbuntu_${VERSION_ID}/ /' > /etc/apt/sources.list.d/devel:kubic:libcontainers:stable.list"
wget -nv https://download.opensuse.org/repositories/devel:kubic:libcontainers:stable/xUbuntu_${VERSION_ID}/Release.key -O- | sudo apt-key add -
sudo apt update
sudo apt -y install podman slirp4netns crun
sudo apt -y install fuse-overlayfs podman slirp4netns crun
fi
- $CONTAINER info
@ -95,4 +95,4 @@ script:
- INSTALL="${DISTRO#${REPO}}"
- INSTALL="${INSTALL%%:*}"
- INSTALL="${INSTALL%%/*}"
- $CONTAINER run $CONTAINER_ARGS -t ima-evm-utils /bin/sh -c "if [ \"$VARIANT\" ]; then ARCH=\"$ARCH\" ./ci/$INSTALL.$VARIANT.sh; fi && ARCH=\"$ARCH\" CC=\"$CC\" TSS=\"$TSS\" ./ci/$INSTALL.sh && if [ "$COMPILE_SSL" ]; then COMPILE_SSL="$COMPILE_SSL" ./tests/install-openssl3.sh; fi && if [ ! \"$VARIANT\" ]; then which tpm_server || which swtpm || if which tssstartup; then ./tests/install-swtpm.sh; fi; fi && CC=\"$CC\" VARIANT=\"$VARIANT\" ./build.sh"
- $CONTAINER run $CONTAINER_ARGS -t ima-evm-utils /bin/sh -c "if [ \"$VARIANT\" ]; then ARCH=\"$ARCH\" ./ci/$INSTALL.$VARIANT.sh; fi && ARCH=\"$ARCH\" CC=\"$CC\" TSS=\"$TSS\" ./ci/$INSTALL.sh && if [ \"$COMPILE_SSL\" ]; then COMPILE_SSL=\"$COMPILE_SSL\" VARIANT=\"$VARIANT\" ./tests/install-openssl3.sh; fi && if [ ! \"$VARIANT\" ]; then which tpm_server || which swtpm || if which tssstartup; then ./tests/install-swtpm.sh; fi; fi && CC=\"$CC\" VARIANT=\"$VARIANT\" COMPILE_SSL=\"$COMPILE_SSL\" ./build.sh"

View File

@ -1,4 +1,8 @@
SUBDIRS = src tests
if HAVE_PANDOC
SUBDIRS += doc
endif
if MANPAGE_DOCBOOK_XSL
dist_man_MANS = evmctl.1
endif

33
NEWS
View File

@ -1,3 +1,32 @@
2023-2-24 Mimi Zohar <zohar@linux.ibm.com>
version 1.5:
* CI changes:
* New: UML kernel testing environment
* Support for running specific test(s)
* Update distros
* Update software release versions
* New features:
* Signing fs-verity signatures
* Reading TPM 2.0 PCRs via sysfs interface
* New tests:
* Missing IMA mmapped file measurements
* Overlapping IMA policy rules
* EVM portable signatures
* fs-verity file measurements in the IMA measurement list
* Build and library changes:
* OpenSSL 3.0 version related changes
* New configuration options: --disable-engine, --enable-sigv1
* Deprecate IMA signature v1 format
* Misc bug fixes and code cleanup:
* memory leaks, bounds checking, use after free
* Fix and update test output
* Add missing sanity checks
* Documentation:
* Store the sourceforge ima-evm-utils wiki for historical
purposes.
2021-10-22 Mimi Zohar <zohar@linux.ibm.com>
version 1.4:
@ -64,7 +93,7 @@
the TPM PCRs, verify the IMA template data digest against the
template data. (Based on LTP "--verify" option.)
- Ignore file measurement violations while verifying the IMA
measurment list. (Based on LTP "--validate" option.)
measurement list. (Based on LTP "--validate" option.)
- Verify the file data signature included in the measurement list
based on the file hash also included in the measurement list
(--verify-sig)
@ -213,7 +242,7 @@
2012-04-02 Dmitry Kasatkin <dmitry.kasatkin@intel.com>
version 0.1.0
* Fully functional version for lastest 3.x kernels
* Fully functional version for latest 3.x kernels
2011-08-24 Dmitry Kasatkin <dmitry.kasatkin@intel.com>

38
README
View File

@ -25,28 +25,34 @@ COMMANDS
--version
help <command>
import [--rsa] pubkey keyring
sign [-r] [--imahash | --imasig ] [--portable] [--key key] [--pass password] file
import [--rsa (deprecated)] pubkey keyring
sign [-r] [--imahash | --imasig ] [--key key] [--pass[=<password>] file
verify file
ima_boot_aggregate [--pcrs hash-algorithm,file] [TPM 1.2 BIOS event log]
ima_sign [--sigfile] [--key key] [--pass password] file
ima_sign [--sigfile] [--key key] [--pass[=<password>]] file
ima_verify file
ima_setxattr [--sigfile file]
ima_hash file
ima_measurement [--ignore-violations] [--verify-sig [--key "key1, key2, ..."]] [--pcrs [hash-algorithm,]file [--pcrs hash-algorithm,file] ...] file
ima_measurement [--ignore-violations] [--verify-sig [--key "key1, key2, ..."]]
[--pcrs [hash-algorithm,]file [--pcrs hash-algorithm,file] ...]
[--verify-bank hash-algorithm] file
ima_boot_aggregate [--pcrs hash-algorithm,file] [TPM 1.2 BIOS event log]
[--hwtpm]
ima_fix [-t fdsxm] path
sign_hash [--key key] [--pass password]
ima_clear [-t fdsxm] path
sign_hash [--veritysig] [--key key] [--pass=<password>]
hmac [--imahash | --imasig ] file
OPTIONS
-------
-a, --hashalgo sha1, sha224, sha256, sha384, sha512
-a, --hashalgo sha1, sha224, sha256, sha384, sha512, streebog256, streebog512 (default: sha256)
-s, --imasig make IMA signature
--veritysig sign an fs-verity file digest hash
-d, --imahash make IMA hash
-f, --sigfile store IMA signature in .sig file instead of xattr
--xattr-user store xattrs in user namespace (for testing purposes)
--rsa use RSA key type and signing scheme v1
--rsa use RSA key type and signing scheme v1 (deprecated)
-k, --key path to signing key (default: /etc/keys/{privkey,pubkey}_evm.pem)
or a pkcs11 URI
--keyid n overwrite signature keyid with a 32-bit value in hex (for signing)
@ -63,7 +69,15 @@ OPTIONS
--smack use extra SMACK xattrs for EVM
--m32 force EVM hmac/signature for 32 bit target system
--m64 force EVM hmac/signature for 64 bit target system
--engine e preload OpenSSL engine e (such as: gost)
--engine e preload OpenSSL engine e (such as: gost) is deprecated
--ino use custom inode for EVM
--uid use custom UID for EVM
--gid use custom GID for EVM
--mode use custom Mode for EVM
--generation use custom Generation for EVM(unspecified: from FS, empty: use 0)
--ima use custom IMA signature for EVM
--selinux use custom Selinux label for EVM
--caps use custom Capabilities for EVM(unspecified: from FS, empty: do not use)
--pcrs file containing TPM pcrs, one per hash-algorithm/bank
--ignore-violations ignore ToMToU measurement violations
--verify-sig verify the file signature based on the file hash, both
@ -205,7 +219,7 @@ asymmetric keys support:
Configuration file x509_evm.genkey:
# Begining of the file
# Beginning of the file
[ req ]
default_bits = 1024
distinguished_name = req_distinguished_name
@ -256,7 +270,7 @@ following steps:
Configuration file ima-local-ca.genkey:
# Begining of the file
# Beginning of the file
[ req ]
default_bits = 2048
distinguished_name = req_distinguished_name
@ -287,7 +301,7 @@ Produce X509 in DER format for using while building the kernel:
Configuration file ima.genkey:
# Begining of the file
# Beginning of the file
[ req ]
default_bits = 1024
distinguished_name = req_distinguished_name

View File

@ -2,7 +2,7 @@
AC_DEFUN([PKG_ARG_ENABLE],
[
AC_MSG_CHECKING(whether to enable $1)
AC_ARG_ENABLE([$1], AC_HELP_STRING([--enable-$1], [enable $1 (default is $2)]),
AC_ARG_ENABLE([$1], AS_HELP_STRING([--enable-$1], [enable $1 (default is $2)]),
[pkg_cv_enable_$1=$enableval],
[AC_CACHE_VAL([pkg_cv_enable_$1], [pkg_cv_enable_$1=$2])])
if test $pkg_cv_enable_$1 = yes; then

View File

@ -1,6 +1,16 @@
#!/bin/sh
# Copyright (c) 2020 Petr Vorel <pvorel@suse.cz>
if [ -n "$CI" ]; then
# If we under CI only thing we can analyze is logs so better to enable
# verbosity to a maximum.
set -x
# This is to make stdout and stderr synchronous in the logs.
exec 2>&1
mount -t securityfs -o rw securityfs /sys/kernel/security
fi
set -e
CC="${CC:-gcc}"
@ -32,6 +42,14 @@ log_exit()
cd `dirname $0`
if [ "$COMPILE_SSL" ]; then
echo "COMPILE_SSL: $COMPILE_SSL"
export CFLAGS="-I/opt/openssl3/include $CFLAGS"
export LD_LIBRARY_PATH="/opt/openssl3/lib64:/opt/openssl3/lib:$HOME/src/ima-evm-utils/src/.libs:$LD_LIBRARY_PATH"
export LDFLAGS="-L/opt/openssl3/lib64 -L/opt/openssl3/lib $LDFLAGS"
export PATH="/opt/openssl3/bin:$HOME/src/ima-evm-utils/src/.libs:$PATH"
fi
case "$VARIANT" in
i386)
echo "32-bit compilation"
@ -79,17 +97,7 @@ VERBOSE=1 make check || ret=$?
title "logs"
if [ $ret -eq 0 ]; then
if [ -f tests/ima_hash.log ]; then
tail -3 tests/ima_hash.log
grep "skipped" tests/ima_hash.log && \
grep "skipped" tests/ima_hash.log | wc -l
fi
if [ -f tests/sign_verify.log ]; then
tail -3 tests/sign_verify.log
grep "skipped" tests/sign_verify.log && \
grep "skipped" tests/sign_verify.log | wc -l
fi
tail -20 tests/boot_aggregate.log
cd tests; make check_logs; cd ..
exit 0
fi

View File

@ -26,9 +26,11 @@ apk add \
attr-dev \
autoconf \
automake \
bash \
diffutils \
docbook-xml \
docbook-xsl \
e2fsprogs-extra \
keyutils-dev \
libtool \
libxslt \
@ -40,9 +42,11 @@ apk add \
pkgconfig \
procps \
sudo \
util-linux \
wget \
which \
xxd
xxd \
gawk
if [ ! "$TSS" ]; then
apk add git

View File

@ -11,7 +11,8 @@ apt-get install -y \
$TSS \
asciidoc \
attr \
docbook-style-xsl \
e2fsprogs \
fsverity-utils-devel \
gnutls-utils \
libattr-devel \
libkeyutils-devel \
@ -21,6 +22,7 @@ apt-get install -y \
openssl-gost-engine \
rpm-build \
softhsm \
util-linux \
wget \
xsltproc \
xxd \

View File

@ -2,6 +2,9 @@
# Copyright (c) 2020 Petr Vorel <pvorel@suse.cz>
set -ex
# workaround for Ubuntu impish asking to interactively configure tzdata
export DEBIAN_FRONTEND="noninteractive"
if [ -z "$CC" ]; then
echo "missing \$CC!" >&2
exit 1
@ -37,6 +40,7 @@ $apt \
debianutils \
docbook-xml \
docbook-xsl \
e2fsprogs \
gzip \
libattr1-dev$ARCH \
libkeyutils-dev$ARCH \
@ -47,8 +51,10 @@ $apt \
pkg-config \
procps \
sudo \
util-linux \
wget \
xsltproc
xsltproc \
gawk
$apt xxd || $apt vim-common
$apt libengine-gost-openssl1.1$ARCH || true

View File

@ -25,9 +25,12 @@ yum -y install \
automake \
diffutils \
docbook-xsl \
e2fsprogs \
git-core \
gnutls-utils \
gzip \
keyutils-libs-devel \
kmod \
libattr-devel \
libtool \
libxslt \
@ -38,9 +41,16 @@ yum -y install \
pkg-config \
procps \
sudo \
util-linux \
vim-common \
wget \
which
which \
zstd \
systemd \
keyutils \
e2fsprogs \
acl \
libcap
yum -y install docbook5-style-xsl || true
yum -y install swtpm || true
@ -50,3 +60,9 @@ if [ -f /etc/centos-release ]; then
yum -y install epel-release
fi
yum -y install softhsm || true
# haveged is available via EPEL on CentOS stream8.
yum -y install haveged || true
./tests/install-fsverity.sh
./tests/install-mount-idmapped.sh

View File

@ -26,6 +26,7 @@ zypper --non-interactive install --force-resolution --no-recommends \
diffutils \
docbook_5 \
docbook5-xsl-stylesheets \
e2fsprogs \
gzip \
ibmswtpm2 \
keyutils-devel \
@ -37,10 +38,12 @@ zypper --non-interactive install --force-resolution --no-recommends \
pkg-config \
procps \
sudo \
util-linux \
vim \
wget \
which \
xsltproc
xsltproc \
gawk
zypper --non-interactive install --force-resolution --no-recommends \
gnutls openssl-engine-libp11 softhsm || true

View File

@ -1,7 +1,7 @@
# autoconf script
AC_PREREQ([2.65])
AC_INIT(ima-evm-utils, 1.4, zohar@linux.ibm.com)
AC_INIT(ima-evm-utils, 1.5, zohar@linux.ibm.com)
AM_INIT_AUTOMAKE([foreign])
AC_CONFIG_HEADERS([config.h])
AC_CONFIG_MACRO_DIR([m4])
@ -15,16 +15,14 @@ AM_PROG_CC_C_O
#AC_PROG_CXX
#AC_PROG_CPP
AC_PROG_INSTALL
AC_PROG_LIBTOOL
#AC_PROG_LN_S
AC_CHECK_PROG(have_pandoc, [pandoc], [yes], [no])
AM_CONDITIONAL([HAVE_PANDOC], [test "x$have_pandoc" = "xyes"])
LT_INIT
# FIXME: Replace `main' with a function in `-lpthread':
#AC_CHECK_LIB([pthread], [main])
# Checks for header files.
AC_HEADER_STDC
PKG_CHECK_MODULES(LIBCRYPTO, [libcrypto >= 0.9.8 ])
AC_SUBST(KERNEL_HEADERS)
AC_CHECK_HEADER(unistd.h)
@ -53,6 +51,16 @@ AC_ARG_ENABLE([openssl_conf],
AC_DEFINE(DISABLE_OPENSSL_CONF, 1, [Define to disable loading of openssl config by evmctl.])
fi], [enable_openssl_conf=yes])
AC_ARG_ENABLE(sigv1,
AS_HELP_STRING([--enable-sigv1], [Build ima-evm-utils with signature v1 support]))
AM_CONDITIONAL([CONFIG_SIGV1], [test "x$enable_sigv1" = "xyes"])
AS_IF([test "$enable_sigv1" != "yes"], [enable_sigv1="no"])
AC_ARG_ENABLE(engine,
[AS_HELP_STRING([--disable-engine], [build ima-evm-utils without OpenSSL engine support])],,[enable_engine=yes])
AC_CHECK_LIB([crypto], [ENGINE_init],, [enable_engine=no])
AM_CONDITIONAL([CONFIG_IMA_EVM_ENGINE], [test "x$enable_engine" = "xyes"])
#debug support - yes for a while
PKG_ARG_ENABLE(debug, "yes", DEBUG, [Enable Debug support])
if test $pkg_cv_enable_debug = yes; then
@ -73,6 +81,8 @@ AX_DEFAULT_HASH_ALGO([$KERNEL_HEADERS])
AC_CONFIG_FILES([Makefile
src/Makefile
tests/Makefile
doc/Makefile
doc/sf/Makefile
packaging/ima-evm-utils.spec
])
AC_OUTPUT
@ -87,5 +97,8 @@ echo " openssl-conf: $enable_openssl_conf"
echo " tss2-esys: $ac_cv_lib_tss2_esys_Esys_Free"
echo " tss2-rc-decode: $ac_cv_lib_tss2_rc_Tss2_RC_Decode"
echo " ibmtss: $ac_cv_header_ibmtss_tss_h"
echo " sigv1: $enable_sigv1"
echo " engine: $enable_engine"
echo " doc: $have_doc"
echo " pandoc: $have_pandoc"
echo

1
doc/Makefile.am Normal file
View File

@ -0,0 +1 @@
SUBDIRS = sf

6
doc/sf/Makefile.am Normal file
View File

@ -0,0 +1,6 @@
noinst_DATA = sf-wiki.html
sf-wiki.html:sf-wiki.md
pandoc $+ -f markdown -t html > $@
CLEANFILES = sf-wiki.html

46
doc/sf/sf-diagram.html Normal file
View File

@ -0,0 +1,46 @@
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML>
</STYLE>
<BODY LANG="en-US" DIR="LTR">
<p style="text-align: center; font-family:arial">
<FONT SIZE=+3><b><a href="https://sourceforge.net/p/linux-ima/wiki/Home">
See documentation at Linux IMA/EVM Wiki</a></b></FONT>
<br>
<FONT SIZE=+3><b>Linux Integrity Subsystem</b></FONT>
<p>The goals of the kernel integrity subsystem are to detect if files have
been accidentally or maliciously altered, both remotely and locally,
appraise a file's measurement against a "good" value stored as an extended
attribute, and enforce local file integrity. These goals are complementary
to Mandatory Access Control(MAC) protections provided by LSM modules, such as
SElinux and Smack, which, depending on policy, can attempt to protect file
integrity. The following modules provide several integrity functions:</p>
<object type="text/html" style="float:right" height=450 data="tcg.html-20100504"></object>
<UL>
<LI><B>Collect</B> - measure a file before it is accessed. </li>
<LI><B>Store</B> - add the measurement to a kernel resident list and, if a
hardware Trusted Platform Module (TPM) is present, extend the IMA PCR </li>
<LI><B>Attest</B> -if present, use the TPM to sign the IMA PCR value, to
allow a remote validation of the measurement list.</li>
<LI><B>Appraise</B> - enforce local validation of a measurement against a
'good' value stored in an extended attribute of the file.</li>
<LI><B>Protect</B> - protect a file's security extended attributes
</UL>
<p>The first three functions were introduced with Integrity Measurement
Architecture (IMA) in 2.6.30. The EVM/IMA-appraisal patches add support for
the last two features.</p>
<p>For additional information about the Linux integrity subsystem, refer to the
<a href="http://sourceforge.net/apps/mediawiki/linux-ima/index.php?title=Main_Page">Wiki</a>.
</p>
<H3><a name="Trusted-Computing">Trusted Computing: architecture and opensource components</a></H3>
<P> IMA measurement, one component of the kernel's integrity subsystem, is part
of an overall Integrity Architecture based on the
<a href="https://www.trustedcomputinggroup.org/home">Trusted Computing Group's
</a> open standards, including Trusted Platform Module (TPM), Trusted Boot,
Trusted Software Stack (TSS), Trusted Network Connect (TNC), and Platform
Trust Services (PTS). The diagram shows how these standards relate, and
provides links to the respective specifications and open source
implementations. IMA and EVM can still run on platforms without a
hardware TPM, although without the hardware guarantee of compromise
detection.
</P>
</BODY></HTML>

99
doc/sf/sf-tcg.html Normal file
View File

@ -0,0 +1,99 @@
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML>
<STYLE TYPE="text/css">
.tcg TD, .tcgcell
{
background-color:lightblue;
color:black;
font-family:sans-serif;
font-weight:700;
padding:0;
}
.tcg2 TD, .tcg2cell
{
background-color:white;
color:black;
font-family:sans-serif;
padding:5;
border:0;
}
</STYLE>
<table CLASS="tcg">
<tr> <th COLSPAN=2 ><HREF="http://www.trustedcomputinggroup.org/trusted_computing"></A></th> </tr>
<tr>
<td><h3>Applications
<table ALIGN=RIGHT CLASS="tcg2">
<tr>
<th>spec</th>
<th>info</th>
</tr>
<tr>
<td><a href="http://www.trustedcomputinggroup.org/resources/infrastructure_work_group_platform_trust_services_interface_specification_version_10" target="_top">PTS</a></td>
<td><a href="http://openpts.sourceforge.jp" target="_top">OpenPTS</a></td>
</tr>
<tr><td style="background-color: lightblue"></td>
<td><a href="http://sourceforge.net/projects/trousers/files/tpm-tools/tpm-tools-1.3.2.tar.gz/download" target="_top">tpm-tools</a></td>
</tr>
</table>
<h2></td>
</tr>
<tr>
<td><h3>Libraries
<table ALIGN=RIGHT CLASS="tcg2">
<tr>
<th>spec</th>
<th>info</th>
</tr>
<tr><td><a href="http://www.trustedcomputinggroup.org/developers/software_stack" target="_top">TSS</a></td>
<td><a href="http://trousers.sourceforge.net/" target="_top">TrouSerS</a></td>
</tr>
</table>
</td>
</tr>
<tr><td><h3>Linux Kernel
<table ALIGN=RIGHT CLASS="tcg2">
<tr>
<th>spec</th>
<th>info</th>
</tr>
<tr><td style="background-color: lightblue"></td>
<td><a href="http://linux-ima.sourceforge.net/#Integrity Measurement Architecture" target="_top">IMA</a>,
<a href="http://linux-ima.sourceforge.net/#Linux Extended Verification Module" target="_top">EVM</a></td>
</td>
</tr>
<td><a href="http://www.trustedcomputinggroup.org/files/resource_files/87BCE22B-1D09-3519-ADEBA772FBF02CBD/TCG_PCClientTPMSpecification_1-20_1-00_FINAL.pdf" target="_top">tpm-1.2</a></td>
<td><a href="http://tpmdd.sourceforge.net" target="_top">TPM driver</a></td>
</tr>
</table>
</td>
</tr>
<tr><td><h3>Boot
<table ALIGN=RIGHT CLASS="tcg2">
<tr>
<th>spec</th>
<th>info</th>
</tr>
<tr><td><a href="http://www.trustedcomputinggroup.org/resources/pc_client_work_group_specific_implementation_specification_for_conventional_bios_specification_version_12" target="_top">BIOS</a></td>
<td><a href="http://sourceforge.jp/projects/openpts/wiki/GRUB-IMA" target="_top">GRUB-IMA</a>,
<a href="http://sourceforge.net/projects/tboot" target="_top"> TBOOT</a></td>
</tr>
</td>
</tr>
</table>
<tr><td><h3>Hardware
<table ALIGN=RIGHT CLASS="tcg2">
<tr>
<th>spec</th>
<th>info</th>
</tr>
<tr>
<td><a href="http://www.trustedcomputinggroup.org/developers/trusted_platform_module" target="_top">TPM</a></td>
<td><a href="http://ibmswtpm.sourceforge.net" target="_top">(swTPM)</a></td>
</tr>
</table>
</td>
</tr>
</table>
</BODY>
</HTML>

932
doc/sf/sf-wiki.md Normal file
View File

@ -0,0 +1,932 @@
The goals of the kernel integrity subsystem are to detect if files have been accidentally or maliciously altered, both remotely and locally, appraise a file's measurement against a "good" value stored as an extended attribute, and enforce local file integrity. These goals are complementary to Mandatory Access Control(MAC) protections provided by LSM modules, such as SElinux and Smack, which, depending on policy, can attempt to protect file integrity.
[TOC]
## Overview
### Features
The following modules provide several integrity functions:
- **Collect** measure a file before it is accessed.
- **Store** add the measurement to a kernel resident list and, if a
hardware Trusted Platform Module (TPM) is present, extend the IMA
PCR
- **Attest** if present, use the TPM to sign the IMA PCR value, to
allow a remote validation of the measurement list.
- **Appraise** enforce local validation of a measurement against a
“good” value stored in an extended attribute of the file.
- **Protect** protect a file's security extended attributes
(including appraisal hash) against off-line attack.
- **Audit** audit the file hashes.
The first three functions were introduced with Integrity Measurement Architecture ([IMA](#integrity-measurement-architecture-ima)) in 2.6.30. The "appraise" and "protect" features were originally posted as a single [EVM](#linux-extended-verification-module-evm)/[IMA-appraisal](#ima-appraisal) patch set for in the 2.6.36 timeframe, but were subsequently split. EVM, the "protect" feature, was upstreamed in Linux 3.2, using a simplier and more secure method for loading the 'evm-key', based on the new Kernel Key Retention [Trusted and Encrypted keys](#creating-trusted-and-evm-encrypted-keys). EVM support for protecting file metadata based on digital signatures was upstreamed in the Linux 3.3. IMA-appraisal, the fourth aspect, appraising a file's integrity, was upstreamed in Linux 3.7.
The goals, design, and benefits of these features are further described in the whitepaper ["An Overview of the Linux Integrity Subsystem"](http://downloads.sf.net/project/linux-ima/linux-ima/Integrity_overview.pdf "http://downloads.sf.net/project/linux-ima/linux-ima/Integrity_overview.pdf").
### Components
IMA-measurement, one component of the kernel's integrity subsystem, is part of an overall Integrity Architecture based on the Trusted Computing Group's open standards, including Trusted Platform Module (TPM), Trusted Boot, Trusted Software Stack (TSS), Trusted Network Connect (TNC), and Platform Trust Services (PTS). The linux-ima project page contains a [diagram](http://linux-ima.sourceforge.net/) showing how these standards relate, and provides links to the respective specifications and open source implementations. IMA-measurement and EVM can still run on platforms without a hardware TPM, although without the hardware guarantee of compromise detection.
IMA-appraisal, a second component of the kernel's integrity subsystem, extends the "secure boot" concept of verifying a file's integrity, before transferring control or allowing the file to be accessed by the OS.
IMA-audit, another component of the kernel's integrity subsystem, includes file hashes in the system audit logs, which can be used to augment existing system security analytics/forensics.
The IMA-measurement, IMA-appraisal, and IMA-audit aspects of the kernel's integrity subsystem complement each other, but can be configured and used independently of each other.
## Integrity Measurement Architecture (IMA-measurement)
IMA-measurement is an open source trusted computing component. IMA maintains a runtime measurement list and, if anchored in a hardware Trusted Platform Module(TPM), an aggregate integrity value over this list. The benefit of anchoring the aggregate integrity value in the TPM is that the measurement list cannot be compromised by any software attack, without being detectable. Hence, on a trusted boot system, IMA-measurement can be used to attest to the system's runtime integrity.
### Enabling IMA-measurement
IMA was first included in the 2.6.30 kernel. For distros that enable IMA by default in their kernels, collecting IMA measurements simply requires rebooting the kernel with a builtin "ima_policy=" on the boot command line. (Fedora/RHEL may also require the boot command line parameter 'ima=on'.)
To determine if your distro enables IMA by default, mount securityfs (mount -t securityfs security /sys/kernel/security), if it isn't already mounted, and then check if '<securityfs>/integrity/ima' exists. If it exists, IMA is indeed enabled. On systems without IMA enabled, [recompile the kernel](#compiling-the-kernel-with-evmima-appraisal-enabled) with the config option 'CONFIG_IMA' enabled.
### Controlling IMA-measurement
IMA is controlled with several kernel command line parameters:
ima_audit= informational audit logging
Format: { "0" | "1" }
0 -- normal integrity auditing messages. (Default)
1 -- enable additional informational integrity auditing messages.
(eg. Although file measurements are only added to the measurement list once and cached, if the inode is flushed, subsequent access to the inode will result in re-measuring the file and attempting to add the measurement again to the measurement list. Enabling ima_audit will log such attempts.)
ima_policy= builtin policy
Format: {"tcb" | "appraise_tcb" | "secure-boot"}
**NEW** Linux-4.13 default: no policy
ima_template= template used
Format: { "ima" | "ima-ng" | "ima-sig" }
Linux 3.13 default: "ima-ng"
ima_hash= hash used
Format: { "sha1" | "md5" | "sha256" | "sha512" | "wp512" | ... }
'ima' template default: "sha1"
Linux 3.13 default: "sha256"
ima_tcb (deprecated)
If specified, enables the TCB policy, which meets the needs of the Trusted Computing Base. This means IMA will measure all programs exec'd, files mmap'd for exec, and all files opened for read by uid=0.
### IMA Measurement List
IMA-measurements maintains a runtime measurement list, which can be displayed as shown below.
- mount securityfs as /sys/kernel/security
$ su -c 'mkdir /sys/kernel/security'
$ su -c 'mount -t securityfs securityfs /sys/kernel/security'
Modify /etc/fstab to mount securityfs on boot.
- display the runtime measurement list (Only root is allowed access to securityfs files.)
Example 1: 'ima-ng' template
$ su -c 'head -5 /sys/kernel/security/ima/ascii_runtime_measurements'
PCR template-hash filedata-hash filename-hint
10 91f34b5c671d73504b274a919661cf80dab1e127 ima-ng sha1:1801e1be3e65ef1eaa5c16617bec8f1274eaf6b3 boot_aggregate
10 8b1683287f61f96e5448f40bdef6df32be86486a ima-ng sha256:efdd249edec97caf9328a4a01baa99b7d660d1afc2e118b69137081c9b689954 /init
10 ed893b1a0bc54ea5cd57014ca0a0f087ce71e4af ima-ng sha256:1fd312aa6e6417a4d8dcdb2693693c81892b3db1a6a449dec8e64e4736a6a524 /usr/lib64/ld-2.16.so
10 9051e8eb6a07a2b10298f4dc2342671854ca432b ima-ng sha256:3d3553312ab91bb95ae7a1620fedcc69793296bdae4e987abc5f8b121efd84b8 /etc/ld.so.cache
PCR: default CONFIG_IMA_MEASURE_PCR_IDX is 10
template-hash: sha1 hash(filedata-hash length, filedata-hash, pathname length, pathname)
filedata-hash: sha256 hash(filedata)
Example 2: 'ima-sig' template (same format as ima-ng, but with an appended signature when present)
PCR template-hash filedata-hash filename-hint file-signature
10 f63c10947347c71ff205ebfde5971009af27b0ba ima-sig sha256:6c118980083bccd259f069c2b3c3f3a2f5302d17a685409786564f4cf05b3939 /usr/lib64/libgspell-1.so.1.0.0 0302046e6c10460100aa43a4b1136f45735669632ad ...
10 595eb9bf805874b459ce073af158378f274ea961 ima-sig sha256:8632769297867a80a9614caa98034d992441e723f0b383ca529faa306c640638 /usr/lib64/gedit/plugins/libmodelines.so 0302046e6c104601002394b70ab93 ...
Example 3: *original* 'ima' template
PCR template-hash filedata-hash filename-hint
10 7971593a7ad22a7cce5b234e4bc5d71b04696af4 ima b5a166c10d153b7cc3e5b4f1eab1f71672b7c524 boot_aggregate
10 2c7020ad8cab6b7419e4973171cb704bdbf52f77 ima e09e048c48301268ff38645f4c006137e42951d0 /init
10 ef7a0aff83dd46603ebd13d1d789445365adb3b3 ima 0f8b3432535d5eab912ad3ba744507e35e3617c1 /init
10 247dba6fc82b346803660382d1973c019243e59f ima 747acb096b906392a62734916e0bb39cef540931 ld-2.9.so
10 341de30a46fa55976b26e55e0e19ad22b5712dcb ima 326045fc3d74d8c8b23ac8ec0a4d03fdacd9618a ld.so.cache
PCR: default CONFIG_IMA_MEASURE_PCR_IDX is 10
template-hash: sha1 hash(filedata-hash, filename-hint)
filedata-hash: sha1 hash(filedata)
The first element in the runtime measurement list, shown above, is the boot_aggregate. The boot_aggregate is a SHA1 hash over tpm registers 0-7, assuming a TPM chip exists, and zeroes, if the TPM chip does not exist.
- display the bios measurement list entries, used in calculating the boot aggregate
$ su -c 'head /sys/kernel/security/tpm0/ascii_bios_measurements'
0 f797cb88c4b07745a129f35ea01b47c6c309cda9 08 [S-CRTM Version]
0 dca68da0707a9a52b24db82def84f26fa463b44d 01 [POST CODE]
0 dd9efa31c88f467c3d21d3b28de4c53b8d55f3bc 01 [POST CODE]
0 dd261ca7511a7daf9e16cb572318e8e5fbd22963 01 [POST CODE]
0 df22cabc0e09aabf938bcb8ff76853dbcaae670d 01 [POST CODE]
0 a0d023a7f94efcdbc8bb95ab415d839bdfd73e9e 01 [POST CODE]
0 38dd128dc93ff91df1291a1c9008dcf251a0ef39 01 [POST CODE]
0 dd261ca7511a7daf9e16cb572318e8e5fbd22963 01 [POST CODE]
0 df22cabc0e09aabf938bcb8ff76853dbcaae670d 01 [POST CODE]
0 a0d023a7f94efcdbc8bb95ab415d839bdfd73e9e 01 [POST CODE]
### Verifying IMA Measurements
The IMA tests programs are part of the [Linux Test Project.](https://github.com/linux-test-project/ltp/wiki)
- Download, compile, and install the standalone version of the IMA LTP test programs in /usr/local/bin.
$ wget -O ltp-ima-standalone-v2.tar.gz http://downloads.sf.net/project/linux-ima/linux-ima/ltp-ima-standalone-v2.tar.gz
$ tar -xvzf ltp-ima-standalone-v2.tar.gz
ima-tests/Makefile
ima-tests/README
ima-tests/ima_boot_aggregate.c
ima-tests/ima_measure.c
ima-tests/ima_mmap.c
ima-tests/ima_sigv2.c
ima-tests/ltp-tst-replacement.c
ima-tests/pkeys.c
ima-tests/rsa_oid.c
ima-tests/config.h
ima-tests/debug.h
ima-tests/hash_info.h
ima-tests/ima_sigv2.h
ima-tests/list.h
ima-tests/pkeys.h
ima-tests/rsa.h
ima-tests/test.h
$ cd ima-tests
$ make
$ su -c 'make install'
- ima_boot_aggregate <tpm_bios file>
Using the TPM's binary bios measurement list, re-calculate the boot aggregate.
$ su -c '/usr/local/bin/ima_boot_aggregate /sys/kernel/security/tpm0/binary_bios_measurements'
000 f797cb88c4b07745a129f35ea01b47c6c309cda9
000 dca68da0707a9a52b24db82def84f26fa463b44d
< snip >
005 6895eb784cdaf843eaad522e639f75d24d4c1ff5
PCR-00: 07274edf7147abda49200100fd668ce2c3a374d7
PCR-01: 48dff4fbf3a34d56a08dfc1504a3a9d707678ff7
PCR-02: 53de584dcef03f6a7dac1a240a835893896f218d
PCR-03: 3a3f780f11a4b49969fcaa80cd6e3957c33b2275
PCR-04: acb44e9dd4594d3f121df2848f572e4d891f0574
PCR-05: df72e880e68a2b52e6b6738bb4244b932e0f1c76
PCR-06: 585e579e48997fee8efd20830c6a841eb353c628
PCR-07: 3a3f780f11a4b49969fcaa80cd6e3957c33b2275
boot_aggregate:b5a166c10d153b7cc3e5b4f1eab1f71672b7c524
and compare the value with the ascii_runtime_measurement list value.
$ su -c 'cat /sys/kernel/security/ima/ascii_runtime_measurements | grep boot_aggregate'
10 7971593a7ad22a7cce5b234e4bc5d71b04696af4 ima b5a166c10d153b7cc3e5b4f1eab1f71672b7c524 boot_aggregate
<br>
- ima_measure <binary_runtime_measurements> \[--validate\] \[--verify\] \[--verbose\]
using the IMA binary measurement list, calculate the PCR aggregate value
$ su -c '/usr/local/bin/ima_measure /sys/kernel/security/ima/binary_runtime_measurements --validate'
PCRAggr (re-calculated): B4 D1 93 D8 FB 31 B4 DD 36 5D DA AD C1 51 AC 84 FA 88 78 1B
and compare it against the PCR value
$ cat /sys/devices/pnp0/00:0a/pcrs | grep PCR-10
PCR-10: B4 D1 93 D8 FB 31 B4 DD 36 5D DA AD C1 51 AC 84 FA 88 78 1B
### IMA re-measuring files
Part of the TCG requirement is that all Trusted Computing Base (TCB) files be measured, and re-measured if the file has changed, before reading/executing the file. IMA detects file changes based on i_version. To re-measure a file after it has changed, the filesystem must support i_version and, if needed, be mounted with i_version (eg. ext3, ext4). Not all filesystems require the explicit mount option. With commit a2a2c3c8580a ("ima: Use i_version only when filesystem supports it") i_version is considered an optimization. If i_version is not enabled, either because the local filesystem does not support it or the filesystem was not mounted with i_version, the file will now always be re-measured, whether or not the file changed, but only new measurements will be added to the measurement list.
- Attempt to mount a filesystem with i_version support.
$ su -c 'mount -o remount,rw,iversion /home'
mount: you must specify the filesystem type
Attempt to remount '/home' with i_version support, shown above, failed. Please install a version of the [util-linux-ng-2.15-rc1](http://www.kernel.org/pub/linux/utils/util-linux-ng/v2.15/ "http://www.kernel.org/pub/linux/utils/util-linux-ng/v2.15/") package or later.
- To automatically mount a filesystem with i_version support, update /etc/fstab.
UUID=blah /home ext3 defaults,iversion
- Mount the root filesystem with i_version.
- For systems with /etc/rc.sysinit, update the mount options
adding 'iversion':
# Remount the root filesystem read-write.
update_boot_stage RCmountfs
if remount_needed ; then
action $"Remounting root filesystem in read-write mode: " mount -n -o remount,rw,iversion /
fi
- For systems using dracut, root 'mount' options can be specified on the boot
command line using 'rootflags'. Add 'rootflags=i_version'. Unlike 'mount',
which expects 'iversion', notice that on the boot command line 'i_version'
contains an underscore.
### Linux-audit support
As of [Linux-audit](http://people.redhat.com/sgrubb/audit/ "http://people.redhat.com/sgrubb/audit/") 2.0, support for integrity auditing messages is available.
### Defining an LSM specific policy
The ima_tcb default measurement policy in linux-2.6.30 measures all system sensitive files - executables, mmapped libraries, and files opened for read by root. These measurements, the measurement list and the aggregate integrity value, can be used to attest to a system's
runtime integrity. Based on these measurements, a remote party can detect whether critical system files have been modified or if malicious software has been executed.
Default policy
dont_measure fsmagic=PROC_SUPER_MAGIC
dont_measure fsmagic=SYSFS_MAGIC
dont_measure fsmagic=DEBUGFS_MAGIC
dont_measure fsmagic=TMPFS_MAGIC
dont_measure fsmagic=SECURITYFS_MAGIC
dont_measure fsmagic=SELINUX_MAGIC
measure func=BPRM_CHECK
measure func=FILE_MMAP mask=MAY_EXEC
< add LSM specific rules here >
measure func=PATH_CHECK mask=MAY_READ uid=0
But not all files opened by root for read, are necessarily part of the Trusted Computing Base (TCB), and therefore do not need to be measured. Linux Security Modules (LSM) maintain file metadata, which can be leveraged to limit the number of files measured.
Examples: adding LSM specific rules
SELinux:
dont_measure obj_type=var_log_t
dont_measure obj_type=auditd_log_t
Smack:
measure subj_user=_ func=INODE_PERM mask=MAY_READ
To replace the default policy 'cat' the custom IMA measurement policy and redirect the output to "< securityfs >/ima/policy". Both dracut and systemd have been modified to load the custom IMA policy. If the IMA policy contains LSM labels, then the LSM policy must be loaded prior to the IMA policy. (eg. if systemd loads the SELinux policy, then systemd must also load the IMA policy.)
systemd commit c8161158 adds support for loading a custom IMA measurement policy. Simply place the custom IMA policy in /etc/ima/ima-policy. systemd will automatically load the custom policy.
dracut commit 0c71fb6 add initramfs support for loading the custom IMA measurement policy. Build and install dracut (git://git.kernel.org/pub/scm/boot/dracut/dracut.git), to load the custom IMA measurement policy(default: /etc/sysconfig/ima-policy).
For more information on defining an LSM specific measurement/appraisal/audit policy, refer to the kernel Documentation/ABI/testing/ima_policy.
## IMA-appraisal
IMA currently maintains an integrity measurement list used for remote attestation. The IMA-appraisal extension adds local integrity validation and enforcement of the measurement against a "good" value stored as an extended attribute 'security.ima'. The initial method for validating 'security.ima' are hashed based, which provides file data integrity, and digital signature based, which in addition to providing file data integrity, provides authenticity.
### Enabling IMA-appraisal
IMA-appraisal was upstreamed in Linux 3.7. For distros that enable IMA-appraisal by default in their kernels, appraising file measurements requires rebooting the kernel first with the boot command line parameters 'ima_appraise_tcb' and ima_appraise='fix' to [label the filesystem](#labeling-the-filesystem-with-securityima-extended-attributes). Once labeled, reboot with just the 'ima_appraise_tcb' boot command line parameter.
Refer to [compiling the kernel](#compiling-the-kernel-with-evmima-appraisal-enabled) for directions on configuring and building a new kernel with IMA-appraisal support enabled.
### Understanding the IMA-appraisal policy
The IMA-appraisal policy extends the measurement policy ABI with two new keywords: appraise/dont_appraise. The default appraise policy appraises all files owned by root. Like the default measurement policy, the default appraisal policy does not appraise pseudo filesystem files (eg. debugfs, tmpfs, securityfs, or selinuxfs.)
Additional rules can be added to the default IMA measurement/appraisal policy, which take advantage of the SELinux labels, for a more fine grained policy. Refer to Documentation/ABI/testing/ima_policy.
### Labeling the filesystem with 'security.ima' extended attributes
A new boot parameter 'ima_appraise=' has been defined in order to label existing file systems with the 'security.ima' extended attribute.
- ima_appraise= appraise integrity measurements\
Format: { "off" | "log" | "fix" } \
off - is a runtime parameter that turns off integrity appraisal verification.
enforce - verifies and enforces runtime file integrity. \[default\]
fix - for non-digitally signed files, updates the 'security.ima' xattr to reflect the existing file hash.
After building a kernel with IMA-appraisal enabled and verified that the filesystems are mounted with [i_version](#ima-re-measuring-files) support, to label the filesystem, reboot with the boot command line options 'ima_appraise_tcb' and 'ima_appraise=fix'. Opening a file owned by root, will cause the 'security.ima' extended attributes to be written. For example, to label the entire filesystem, execute:
`find / \\( -fstype rootfs -o ext4 -type f \\) -uid 0 -exec head -n 1
'{}' >/dev/null \\;`
### Labeling 'immutable' files with digital signatures
'Immutable' files, such as ELF executables, can be digitally signed, storing the digital signature in the 'security.ima' xattr. Creating the digital signature requires generating an RSA private/public key pair. The private key is used to sign the file, while the public key is used to verify the signature. For example, to digitally sign all kernel modules, replace <RSA private key>, below, with the pathname to your RSA private key, and execute:
`find /lib/modules -name "\*.ko" -type f -uid 0 -exec evmctl sign --imasig '{}' <RSA private key> \;`
evmctl manual page is here [evmctl.1.html](http://linux-ima.sourceforge.net/evmctl.1.html)
### Running with IMA-appraisal
Once the filesystem has been properly labeled, before rebooting, re-install the new labeled kernel. Modify the [initramfs](#building-an-initramfs-to-load-keys) to load the RSA public key on the IMA keyring, using evmctl. Reboot with the 'ima_appraise_tcb' and, possibly, the 'rootflags=i_version' options.
## Extending trusted and secure boot to the OS
( Place holder )
### Including file signatures in the measurement list
The 'ima-sig' template, in addition to the file data hash and the full pathname, includes the file signature, as stored in the 'security.ima' extended attribute.
10 d27747646f317e3ca1205287d0615073fe676bc6 ima-sig sha1:08f8f20c14e89da468bb238
d2012c9458ae67f6a /usr/bin/mkdir 030202afab451100802b22e3ed9f6a70fb5babf030d1181
8152b493bd6bfd916005fad7fdcfd7f88d43f6cffaf6fd1ea3b75032dd702b661d4717729e4a3fa4
ee95a47f239955491fc8064eca8cb96302d305d59750ae4ffde0a5f615f910475eee72ae0306e4ae
0269d7d04af2a485898eec3286795d621e83b7dedc99f5019b7ee49b189f3ded0a2
# getfattr -m ^security --dump -e hex /usr/bin/mkdir
# file: usr/bin/mkdir
security.evm=0x0238b0cdd9e97d5bed3bcde5a4793ef8da6fe7c7cc
security.ima=0x030202afab451100802b22e3ed9f6a70fb5babf030d11818152b493bd6bfd916005fad
7fdcfd7f88d43f6cffaf6fd1ea3b75032dd702b661d4717729e4a3fa4ee95a47f239955491fc8064eca8cb
96302d305d59750ae4ffde0a5f615f910475eee72ae0306e4ae0269d7d04af2a485898eec3286795d621e8
3b7dedc99f5019b7ee49b189f3ded0a2
### Signing IMA-appraisal keys
( Place holder )
## IMA-audit
IMA-audit includes file hashes in the audit log, which can be used to augment existing system security analytics/forensics. IMA-audit extends the IMA policy ABI with the policy action keyword - "audit".
Example policy to audit executable file hashes
audit func=BPRM_CHECK
## Linux Extended Verification Module (EVM)
EVM detects offline tampering of the security extended attributes (e.g. security.selinux, security.SMACK64, security.ima), which is the basis for LSM permission decisions and, with the IMA-appraisal extension, integrity appraisal decisions. EVM provides a framework, and two methods for detecting offline tampering of the security extended attributes. The initial method maintains an HMAC-sha1 across a set of security extended attributes, storing the HMAC as the extended attribute 'security.evm'. The other method is based on a digital signature of the security extended attributes hash. To verify the integrity of an extended attribute, EVM exports evm_verifyxattr(), which re-calculates either the HMAC or the hash, and compares it with the version stored in 'security.evm'.
### Enabling EVM
EVM was upstreamed in Linux 3.2. EVM-digital-signatures is currently in the Linux 3.3 release candidate.
Refer to [compiling the kernel](#compiling-the-kernel-with-evmima-appraisal-enabled), for directions on configuring and building a new kernel with EVM support.
### Running EVM
EVM is configured automatically to protect standard “security” extended attributes:
- security.ima (IMA's stored “good” hash for the file)
- security.selinux (the selinux label/context on the file)
- security.SMACK64 (Smack's label on the file)
- security.capability (Capability's label on executables)
EVM protects the configured extended attributes with an HMAC across their data, keyed with an EVM key provided at boot time. EVM looks for this key named 'evm-key' on root's key ring. Refer to [trusted and EVM encrypted keys](#creating-trusted-and-evm-encrypted-keys), for directions on creating EVM keys. Once loaded, EVM can be activated by writing a '1' to the evm securityfs file: `**echo "1" >/sys/kernel/security/evm**`
Before EVM is activated, any requested integrity appraisals are unknown, so the EVM startup should be done early in the boot process, preferably entirely within the kernel and initramfs (which are measured by trusted grub) and before any reference to the real root filesystem. To build an initramfs with EVM enabled, build and install dracut (git://git.kernel.org/pub/scm/boot/dracut/dracut.git), which contains the trusted and EVM dracut modules.
### Labeling the filesystem with 'security.evm'
A new boot parameter 'evm=fix' has been defined in order to label existing file systems with the 'security.evm' extended attribute.
After building a kernel with EVM, IMA-appraisal, and trusted and encrypted keys enabled, installed the trusted and EVM dracut modules, created the EVM key, and verified that the filesystems are mounted, including root, with [i_version](#ima-re-measuring-files) support, to label the filesystem, reboot with the command line options 'ima_tcb', 'ima_appraise_tcb', 'ima_appraise=fix', 'evm=fix' and, possibly, 'rootflags=i_version'.
Once EVM is started, as existing file metadata changes or as new files are created, EVM assumes that the LSM has approved such changes, and automatically updates the HMACs accordingly, assuming the existing value is valid. In fix mode, opening a file owned by root, will fix the 'security.ima' extended attribute, causing the 'security.evm' extended attribute to be written as well, regardless if the existing security 'ima' or 'evm' extended attributes are valid. To label the entire filesystem, execute:
`find / -fstype ext4 -type f -uid 0 -exec head -n 1 '{}' >/dev/null \;`
The following sign_file script can be used to label all 'ELF' files with EVM and IMA digital signatures, and all other files with just an EVM digital signature.
sign_file:
#!/bin/sh
#label "immutable" files with EVM/IMA digital signatures
#label everything else with just EVM digital signatures
file $1 | grep 'ELF' > /dev/null
if [ $? -eq 0 ]; then
evmctl sign --imasig $1 /home/zohar/privkey_evm.pem
else
evmctl sign --imahash $1 /home/zohar/privkey_evm.pem
fi
Instead of opening the file using head, digitally sign the files:
`find / \( -fstype rootfs -o -fstype ext3 -o -fstype ext4 \) -type f -exec sign_file.sh {} \;`
Once the filesystem has been properly labeled, before rebooting, re-install the new labeled kernel. Modify the initramfs to load the RSA public keys on the EVM and IMA keyring. Reboot with just the 'ima_tcb', 'ima_appraise_tcb' and, possibly, 'rootflags=i_version' options.
## Compiling the kernel with EVM/IMA-appraisal enabled
For those unfamiliar with building a linux kernel, here is a short list of existing websites.
- [http://kernelnewbies.org/KernelBuild](http://kernelnewbies.org/KernelBuild "http://kernelnewbies.org/KernelBuild")
- [http://fedoraproject.org/wiki/BuildingUpstreamKernel](http://fedoraproject.org/wiki/BuildingUpstreamKernel "http://fedoraproject.org/wiki/BuildingUpstreamKernel")
- [https://wiki.ubuntu.com/KernelTeam/GitKernelBuild](https://wiki.ubuntu.com/KernelTeam/GitKernelBuild "https://wiki.ubuntu.com/KernelTeam/GitKernelBuild")
### Configuring the kernel
Depending on the distro, some of these options might already be enabled, but not necessarily as builtin. For distros with recent kernels, download the distro's kernel source and recompile the kernel with the additional .config options, below. (Refer to the distro's documentation for building and installing the kernel from source.)
For IMA, enable the following .config options:
CONFIG_INTEGRITY=y
CONFIG_IMA=y
CONFIG_IMA_MEASURE_PCR_IDX=10
CONFIG_IMA_AUDIT=y
CONFIG_IMA_LSM_RULES=y
For IMA-appraisal, enable the following .config options:
CONFIG_INTEGRITY_SIGNATURE=y
CONFIG_INTEGRITY=y
CONFIG_IMA_APPRAISE=y
EVM has a dependency on encrypted keys, which should be encrypted/decrypted using a trusted key. For those systems without a TPM, the EVM key could be encrypted/decrypted with a user-defined key instead. For EVM, enable the following .config options:
CONFIG_TCG_TPM=y
CONFIG_KEYS=y
CONFIG_TRUSTED_KEYS=y
CONFIG_ENCRYPTED_KEYS=y
CONFIG_INTEGRITY_SIGNATURE=y
CONFIG_INTEGRITY=y
CONFIG_EVM=y
For the new 'ima-ng'/'ima-sig' template support(linux 3.13), clone the stable tree.
$ cd ~/src/kernel
$ git clone git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable.git
$ cd linux-stable
$ git remote update
$ git checkout --track -b linux-3.13.y origin/linux-3.13.y
and enable these additional .config options:
CONFIG_IMA_NG_TEMPLATE=y
CONFIG_IMA_DEFAULT_TEMPLATE="ima-ng"
CONFIG_IMA_DEFAULT_HASH_SHA256=y
### Installing the new kernel
If enabling EVM, before installing the new kernel, follow the directions for creating the EVM encrypted key (#creating_trusted_and_evm_encrypted keys) and EVM/IMA public keys (#creating_and_loading_the_evm_and_ima_publicprivate_keypairs).
Install the kernel as normal.
$ su -c "make modules_install install"
## Creating trusted and EVM encrypted keys
Trusted and encrypted keys are two new key types (upstreamed in 2.6.38) added to the existing kernel key ring service. Both of these new types are variable length symmetic keys and, in both cases, are created in the kernel. User space sees, stores, and loads only encrypted blobs. Trusted Keys require the availability of a Trusted Platform Module (TPM) chip for greater security, while encrypted keys can be used on any system. All user level blobs, are displayed and loaded in hex ascii for convenience, and are integrity verified.
Depending on the distro, trusted and encrypted keys might not be enabled. Refer to [compiling the kernel](#compiling-the-kernel-with-evmima-appraisal_enabled), for directions on configuring and building a new kernel with trusted and encrypted key support.
The trusted and EVM dracut modules, by default, looks for the trusted and EVM encrypted keys in /etc/keys. To create and save the kernel master and EVM keys,
$ su -c 'mkdir -p /etc/keys'
# To create and save the kernel master key (trusted type):
$ su -c 'modprobe trusted encrypted'
$ su -c 'keyctl add trusted kmk-trusted "new 32" @u'
$ su -c 'keyctl pipe `keyctl search @u trusted kmk-trusted` >/etc/keys/kmk-trusted.blob'
# Create the EVM encrypted key
$ su -c 'keyctl add encrypted evm-key "new trusted:kmk-trusted 32" @u'
$ su -c 'keyctl pipe `keyctl search @u encrypted evm-key` >/etc/keys/evm-trusted.blob'
For those systems which don't have a TPM, but want to experiment with EVM, create a user key of 32 random bytes, and an EVM user encrypted key. Unlike trusted/encrypted keys, user type key data is visible to userspace.
$ su -c 'mkdir -p /etc/keys'
# To create and save the kernel master key (user type):
$ su -c 'modprobe trusted encrypted'
$ su -c 'keyctl add user kmk-user "`dd if=/dev/urandom bs=1 count=32 2>/dev/null`" @u'
$ su -c 'keyctl pipe `keyctl search @u user kmk-user` > /etc/keys/kmk-user.blob'
# Create the EVM encrypted key
$ su -c 'keyctl add encrypted evm-key "new user:kmk-user 32" @u'
$ su -c 'keyctl pipe `keyctl search @u encrypted evm-key` >/etc/keys/evm-user.blob'
Update /etc/sysconfig/masterkey to reflect using a 'user-defined' master key type.
MULTIKERNELMODE="NO"
MASTERKEYTYPE="user"
MASTERKEY="/etc/keys/kmk-${MASTERKEYTYPE}.blob"
Similarly update /etc/sysconfig/evm or on the boot command line specify the EVM key filename (eg. 'evmkey=/etc/keys/evm-user.blob'.)
<br>
## Creating and loading the EVM and IMA public/private keypairs
### Digital Signatures: generating an RSA public/private key pair
# generate unencrypted private key
openssl genrsa -out privkey_evm.pem 1024
# or generate encrypted (password protected) private key
openssl genrsa -des3 -out privkey_evm.pem 1024
# or convert unencrypted key to encrypted on
openssl rsa -in /etc/keys/privkey_evm.pem -out privkey_evm_enc.pem -des3
or
openssl pkcs8 -topk8 -in /etc/keys/privkey_evm.pem -out privkey_evm_enc.pem
openssl rsa -pubout -in privkey_evm.pem -out pubkey_evm.pem
### ima-evm-utils: installing the package from source
ima-evem-utils is used to sign files, using the private key, and to load the public keys on the ima/evm keyrings. ima-evm-utils can be cloned from git repo with the following command:
git clone git://linux-ima.git.sourceforge.net/gitroot/linux-ima/ima-evm-utils.git
cd ima-evm-utils
./autogen.sh
./configure
make
sudo make install
evmctl manual page is here [evmctl.1.html](http://linux-ima.sourceforge.net/evmctl.1.html)
### IMA/EVM keyrings: loading the public keys
ima_id=`keyctl newring _ima @u`
evmctl import /etc/keys/pubkey_ima.pem $ima_id
evm_id=`keyctl newring _evm @u`
evmctl import /etc/keys/pubkey_evm.pem $evm_id
## Building an initramfs to load keys
Modify the initramfs to load the EVM encrypted key and the EVM/IMA public keys on their respective keyrings.
### dracut
Dracut commits 0c71fb6 and e1ed2a2 add support for loading the masterkey and the EVM encrypted key, not the EVM/IMA public keys (todo).
0c71fb6 dracut: added new module integrityy
e1ed2a2 dracut: added new module masterkey
Clone dracut (git://git.kernel.org/pub/scm/boot/dracut/dracut.git). By default, the masterkey and integrity modules are not enabled in the dracut git tree. Edit module-setup in both directories, changing the check() return value to 0. 'make' and 'install' dracut.
Create an initramfs:
# dracut -H -f /boot/initramfs-<kernel> <kernel> -M
And add a grub2 menu entry:
# grub2-mkconfig -o /boot/grub2/grub.cfg
### initramfs-tools
To enable IMA/EVM in initramfs-tools it is necessary to add just 2 files to /etc/initramfs-tools directory.
/etc/initramfs-tools/hooks/ima.sh:
#!/bin/sh
echo "Adding IMA binaries"
. /usr/share/initramfs-tools/hook-functions
copy_exec /etc/keys/evm-key
copy_exec /etc/keys/pubkey_evm.pem
copy_exec /etc/ima_policy
copy_exec /bin/keyctl
copy_exec /usr/bin/evmctl /bin/evmctl
/etc/initramfs-tools/scripts/local-top/ima.sh:
#!/bin/sh -e
PREREQ=""
# Output pre-requisites
prereqs()
{
echo "$PREREQ"
}
case "$1" in
prereqs)
prereqs
exit 0
;;
esac
grep -q "ima=off" /proc/cmdline && exit 1
mount -n -t securityfs securityfs /sys/kernel/security
IMA_POLICY=/sys/kernel/security/ima/policy
LSM_POLICY=/etc/ima_policy
grep -v "^#" $LSM_POLICY >$IMA_POLICY
# import EVM HMAC key
keyctl show |grep -q kmk || keyctl add user kmk "testing123" @u
keyctl add encrypted evm-key "load `cat /etc/keys/evm-key`" @u
#keyctl revoke kmk
# import Module public key
mod_id=`keyctl newring _module @u`
evmctl import /etc/keys/pubkey_evm.pem $mod_id
# import IMA public key
ima_id=`keyctl newring _ima @u`
evmctl import /etc/keys/pubkey_evm.pem $ima_id
# import EVM public key
evm_id=`keyctl newring _evm @u`
evmctl import /etc/keys/pubkey_evm.pem $evm_id
# enable EVM
echo "1" > /sys/kernel/security/evm
# enable module checking
#echo "1" > /sys/kernel/security/module_check
generate new initramfs:
update-initramfs -k 3.4.0-rc5-kds+ -u
Edit GRUB bootloader /boot/grub/custom.cfg:
menuentry 'IMA' {
set gfxpayload=$linux_gfx_mode
insmod gzio
insmod part_msdos
insmod ext2
set root='(hd0,msdos1)'
# add following string to kernel command line to enable "fix" mode: "ima_appraise=fix evm=fix"
linux /boot/vmlinuz-3.4.0-rc5-kds+ root=/dev/sda1 ro nosplash ima_audit=1 ima_tcb=1 ima_appraise_tcb=1
initrd /boot/initrd.img-3.4.0-rc5-kds+
}
## IMA policy examples
### Builtin policys
**Enabled on the boot command line:**
*ima_tcb* - measures all files read as root and all files executed
*ima_appraise_tcb* - appraises all files owned by root
### audit log all executables
# audit log all executables
audit func=BPRM_CHECK mask=MAY_EXEC
### Measure nothing, appraise everything
#
# Integrity measure policy
#
# Do not measure anything, but appraise everything
#
# PROC_SUPER_MAGIC
dont_appraise fsmagic=0x9fa0
# SYSFS_MAGIC
dont_appraise fsmagic=0x62656572
# DEBUGFS_MAGIC
dont_appraise fsmagic=0x64626720
# TMPFS_MAGIC
dont_appraise fsmagic=0x01021994
# RAMFS_MAGIC
dont_appraise fsmagic=0x858458f6
# DEVPTS_SUPER_MAGIC
dont_appraise fsmagic=0x1cd1
# BIFMT
dont_appraise fsmagic=0x42494e4d
# SECURITYFS_MAGIC
dont_appraise fsmagic=0x73636673
# SELINUXFS_MAGIC
dont_appraise fsmagic=0xf97cff8c
appraise
## ima-evm-utils
ima-evm-utils package provides the *evmctl* utility that can be used for producing and verifying digital signatures, which are used by Linux kernel integrity subsystem. It can be also used to import keys into the kernel keyring.
evmctl manual page is located here: [http://linux-ima.sourceforge.net/evmctl.1.html](http://linux-ima.sourceforge.net/evmctl.1.html)
<br>
## Using IMA/EVM on Android
Enabling IMA/EVM is not very difficult task but involves few tricky steps related to file system creation and labeling.
Android source code is kept in GIT repositories and usually downloaded using 'repo' tool.
IMA/EVM support was implemented using Android 5.0.2 source tree and tested on Huawei P8.
Set of patches is located [here](https://sourceforge.net/projects/linux-ima/files/Android%20patches/).
### Kernel configuration
Kernel source code is usually located in the 'kernel' folder in the root of the Android source tree.
Huawei P8 runs on HiSilicon Kirin 930/935 64 bit ARM CPU.
Default kernel configuration file is 'kernel/arch/arm64/configs/hisi_3635_defconfig'
Following lines were added:
# Integrity
CONFIG_INTEGRITY=y
CONFIG_IMA=y
CONFIG_IMA_MEASURE_PCR_IDX=10
CONFIG_IMA_AUDIT=y
CONFIG_IMA_LSM_RULES=y
CONFIG_INTEGRITY_SIGNATURE=y
CONFIG_INTEGRITY_ASYMMETRIC_KEYS=y
CONFIG_IMA_APPRAISE=y
CONFIG_EVM=y
# Keys
CONFIG_KEYS=y
CONFIG_KEYS_DEBUG_PROC_KEYS=y
CONFIG_TRUSTED_KEYS=y
CONFIG_ENCRYPTED_KEYS=y
### Kernel command line parameters
Kernel command line parameters are usually specified in board configuration files, such as BoardConfig.mk, for example, 'device/hisi/hi3635/BoardConfig.mk
Add following lines to the file:
BOARD_KERNEL_CMDLINE += ima_audit=1
BOARD_KERNEL_CMDLINE += ima_tcb ima_appraise_tcb
# enable fix mode while testing
BOARD_KERNEL_CMDLINE += ima_appraise=fix evm=fix
### IMA boot initialization
To boot Android, devices usually have boot partition which is flashed with boot.img.
boot.img consist of the kernel and compressed ramdisk which includes Android root filesystem.
boot.img is usually protected using digital signature which is verified by the Android bootloader as a part of Secure Boot process.
Root filesystem contains Android 'init' system and minimal set of tools, which is required to initialize and mount rest of filesystems, including '/system' and '/data'.
Android uses own 'init' system (system/core/init) which reads configuration from '/init.rc' and multiple sourced '/init.*.rc' scripts located in the root folder.
We used to use shell scripts to load IMA/EVM keys and policy. On desktop systems there is no limitation on ramdisk size, but on Android devices it is limited by the size of the boot partition. Android ramdisk/root filesystem does not include shell, but including adding shell, keyctl, evmctl makes ramdisk so big so that boot.img does not fit to the boot partition.
For that reason it was necessary to implement IMA/EVM initialization functionality as native program 'ima-init'.
This patch ([0004-ima_init-tool-to-load-IMA-EVM-keys-and-policy.patch](http://sourceforge.net/projects/linux-ima/files/Android%20patches/0004-ima_init-tool-to-load-IMA-EVM-keys-and-policy.patch/view)) adds 'system/extras/ima-init' project to the Android source tree. It builds '/ima-init' initialization program and generates private and public keys to sign filesystem image usign EVM signatures and verify them during runtime.
ima-init project also includes 'ima_key_gen.sh' script to generate keys and certificates and also basic 'ima_policy', which needs to be changed based on the particular need.
ima-init and public keys are included in the ramdisk root filesystem.
In order to initialize IMA/EVM it is necessary add like following configuration to relevant init.rc file:
service ima /sbin/ima_init
class main
user root
group root
disabled
seclabel u:r:init:s0
oneshot
Above example add 'ima' service which is used to initialize IMA.
IMA service needs to be started using 'start ima' before mounting any real filesystem. For example it was added to the 'on fs' target before mounting 'system' partition.
on fs
mount securityfs none /sys/kernel/security
start ima
wait /dev/block/mmcblk0p38
mount ext4 /dev/block/mmcblk0p38 /system ro
wait /dev/block/mmcblk0p40
mount ext4 /dev/block/mmcblk0p40 /data nosuid nodev noatime data=ordered,i_version
### Mounting filesystems (with iversion)
In order IMA would update 'security.ima' when file changes, it is necessary to mount filesystems with i_version support. Android usually mounts all filesystems in init.rc scripts using 'mount' command. Notice in the example above that '/data' partition is mounted using 'i_version' options.
Desktop mount tool from mount package recognizes iversion option and pass necessary flag to mount system call. Unrecognized options are passed as a string in the last argument of the mount system call to the kernel filesystem module. Kernel filesystem modules recognize 'i_version' option instead of 'iversion'. Thus on the desktop systems it is possible to use both iversion and i_version options.
Android tools do not recognize 'iversion' option. It is necessary to use 'i_version' option.
init.rc 'mount' command options are located after the mount point. All except last are 'init' builtin options and *only* the last option is passed as a string to the mount system call. Thus it is necessary to put 'i_version' option as a last option or to add it to the comma separated option list as above.
### Filesystem labeling
Filesystem labeling with digital signatures has to be done during image creation process. It can be done using two approaches.
The easiest approach is to label ready image. It requires following steps:
1. convert sparse image to normal image using simg2img tool
1. 'loop mount' the image
1. label filesystem using evmctl tool
1. unmount image
1. convert image back to sparse image using img2simg tool
But mount operation would require root privileges to mount filesystem.
Android 'make_ext4fs' tool is used to create filesystem image. It provides support for labeling filesystem using 'security labels' (SELinux). We extended make_ext4fs to compute and set IMA/EVM signatures while creating a filesystem. It uses extended version of 'evmctl' to compute signatures by passing all relevant file metadata using evmctl command line parameters.
Here is a patch that adds IMA/EVM support to the make_ext4fs ([0003-IMA-EVM-labelling-support.patch](http://sourceforge.net/projects/linux-ima/files/Android%20patches/0003-IMA-EVM-labelling-support.patch/view)).
### Additional tools
It is convenient for testing and debugging to have additional tools such as keyctl and getfattr tools on the device.
#### evmctl
For Android, 'evmct' is a host only tool to compute IMA/EVM signatures and convert RSA keys to the kernel binary format.
'evmctl' was extended to pass file metadata using command line parameters:
--ino use custom inode for EVM
--uid use custom UID for EVM
--gid use custom GID for EVM
--mode use custom Mode for EVM
--generation use custom Generation for EVM(unspecified: from FS, empty: use 0)
--ima use custom IMA signature for EVM
--selinux use custom Selinux label for EVM
--caps use custom Capabilities for EVM(unspecified: from FS, empty: do not use)
#### keyctl
This patch ([0002-keyctl-tool.patch](http://sourceforge.net/projects/linux-ima/files/Android%20patches/0002-keyctl-tool.patch/view)) adds project system/extras/keyctl.
#### getfattr
This patch ([0001-getfattr-tool.patch](http://sourceforge.net/projects/linux-ima/files/Android%20patches/0001-getfattr-tool.patch/view)) adds project system/extras/getfattr.
<br>
## Frequently asked questions
- Why is the first entry in the IMA measurement list (/sys/kernel/security ima/ascii_runtime_measurements) are 0's?
The first entry is the TPM boot aggregate containing PCR values 0 -
7. Enable the TPM in BIOS and take ownership.
- How do I take ownership of the TPM?
To take ownership of the TPM, download the tpm-tools, start tcsd (eg. 'service tcsd start'), and execute "tpm_takeownership -u -z". This will set the SRK key to the well-known secret(20 zeroes) and prompt for the TPM owner password.
- Why are there 0x00 entries in the measurement list?
The measurement list is invalidated, when a regular file is opened for read and, at the same time, opened for write. In the majority of cases, these files should not have been measured in the first place (eg. log files). In other cases, the application needs to be fixed.
- Why aren't files re-measured and added to the IMA measurement list
after being updated?
To detect files changing, the filesystem needs to be mounted with i_version support. For the root filesystem, either update /etc/rc.sysinit or add 'rootflags=i_version' boot command line option. For all other filesystems, modify /etc/fstab.
- Why doesn't the measurement list verify?
On some systems, after a suspend/resume, the TPM measurement list does not verify. On those systems, add the boot command line option "tpm.suspend_pcr=< unused PCR >".
- Why are there two /init entries in the measurement list?
The first '/init' is from the initramfs. The second /init is from the root filesystem (eg. /sbin/init). The IMA ng/nglong template patches will provide additional metadata to help correlate measurement entries and files.
- Why am I unable to boot the new EVM/IMA-appraisal enabled kernel?
After building a new kernel with EVM/IMA-appraisal enabled, the filesystem must be labeled with 'security.evm' and 'security.ima' extended attributes. After creating an [EVM
key](#creating_trusted_and_evm_encrypted_keys), boot the new kernel with the 'ima_tcb', 'evm=fix', 'ima_appraise_tcb', 'ima_appraise=fix', and, possibly, 'rootflags=i_version' boot
command line options. Refer to [labeling the filesystem](#labeling-the-filesystem-with-securityima-extended-attributes) with 'security.evm'.
- How do I enable the measurement policy for local/remote attestation, without enabling IMA-appraisal?
Boot with the 'ima_tcb' command line option.
- How do I enable the appraise policy, without the measurement policy?
Boot with the 'ima_appraise_tcb' command line option.
## Links
- IMA/EVM utils man page:
[http://linux-ima.sourceforge.net/evmctl.1.html](http://linux-ima.sourceforge.net/evmctl.1.html)
- Linux IMA project page:
[https://sourceforge.net/projects/linux-ima/](https://sourceforge.net/projects/linux-ima/ "https://sourceforge.net/projects/linux-ima/")
- Old web site:
[http://linux-ima.sourceforge.net/](http://linux-ima.sourceforge.net/ "http://linux-ima.sourceforge.net/")
- GIT repositories:
[https://sourceforge.net/p/linux-ima/ima-evm-utils](https://sourceforge.net/p/linux-ima/ima-evm-utils/)
[Old](/apps/mediawiki/linux-ima/index.php?title=Main_Page_OLD "Old")
Converted from http://sourceforge.net/apps/mediawiki/linux-ima/index.php?title=Main_Page_OLD
[[project_screenshots]]
[[project_admins]]
[[download_button]]

213
kernel-configs/base Normal file
View File

@ -0,0 +1,213 @@
CONFIG_LOCALVERSION="-dont-use"
CONFIG_WATCH_QUEUE=y
CONFIG_AUDIT=y
CONFIG_AUDITSYSCALL=y
CONFIG_HZ_PERIODIC=y
CONFIG_LOG_BUF_SHIFT=17
CONFIG_USER_NS=y
CONFIG_PID_NS=y
CONFIG_CC_OPTIMIZE_FOR_PERFORMANCE=y
CONFIG_KALLSYMS_ALL=y
CONFIG_SYSTEM_DATA_VERIFICATION=y
CONFIG_TRACEPOINTS=y
CONFIG_CON_CHAN="xterm"
CONFIG_SSL_CHAN="pty"
CONFIG_MODULE_SIG_FORMAT=y
CONFIG_MODULE_SIG=y
CONFIG_MODULE_SIG_FORCE=y
CONFIG_MODULE_SIG_ALL=y
CONFIG_MODULE_SIG_SHA1=y
CONFIG_MODULE_SIG_HASH="sha1"
CONFIG_MODULES_TREE_LOOKUP=y
CONFIG_BLK_DEBUG_FS=y
CONFIG_ASN1=y
CONFIG_UNINLINE_SPIN_UNLOCK=y
CONFIG_SLUB=y
CONFIG_COMPACTION=y
CONFIG_COMPACT_UNEVICTABLE_DEFAULT=1
CONFIG_MIGRATION=y
CONFIG_BLK_DEV_LOOP=y
CONFIG_LEGACY_PTY_COUNT=256
CONFIG_NULL_TTY=y
CONFIG_SERIAL_DEV_BUS=y
CONFIG_SERIAL_DEV_CTRL_TTYPORT=y
CONFIG_VALIDATE_FS_PARSER=y
CONFIG_EXT4_FS_POSIX_ACL=y
CONFIG_EXT4_FS_SECURITY=y
CONFIG_EXT4_DEBUG=y
CONFIG_REISERFS_FS_XATTR=y
CONFIG_REISERFS_FS_POSIX_ACL=y
CONFIG_REISERFS_FS_SECURITY=y
CONFIG_FS_POSIX_ACL=y
CONFIG_FS_VERITY=y
CONFIG_FS_VERITY_BUILTIN_SIGNATURES=y
CONFIG_TMPFS_POSIX_ACL=y
CONFIG_TMPFS_XATTR=y
CONFIG_CONFIGFS_FS=y
CONFIG_KEYS=y
CONFIG_ENCRYPTED_KEYS=y
CONFIG_SECURITY=y
CONFIG_SECURITYFS=y
CONFIG_SECURITY_NETWORK=y
CONFIG_SECURITY_PATH=y
CONFIG_LSM="lockdown,yama,loadpin,safesetid,integrity,bpf"
CONFIG_CRYPTO_AEAD2=y
CONFIG_CRYPTO_SKCIPHER=y
CONFIG_CRYPTO_SKCIPHER2=y
CONFIG_CRYPTO_RNG=y
CONFIG_CRYPTO_RNG2=y
CONFIG_CRYPTO_RNG_DEFAULT=y
CONFIG_CRYPTO_AKCIPHER2=y
CONFIG_CRYPTO_AKCIPHER=y
CONFIG_CRYPTO_KPP2=y
CONFIG_CRYPTO_ACOMP2=y
CONFIG_CRYPTO_MANAGER=y
CONFIG_CRYPTO_MANAGER2=y
CONFIG_CRYPTO_NULL2=y
CONFIG_CRYPTO_RSA=y
CONFIG_CRYPTO_ECC=y
CONFIG_CRYPTO_ECDSA=y
CONFIG_CRYPTO_AES=y
CONFIG_CRYPTO_CBC=y
CONFIG_CRYPTO_HMAC=y
CONFIG_CRYPTO_MD5=y
CONFIG_CRYPTO_SHA1=y
CONFIG_CRYPTO_SHA256=y
CONFIG_CRYPTO_SHA512=y
CONFIG_CRYPTO_WP512=y
CONFIG_CRYPTO_LZO=y
CONFIG_CRYPTO_ZSTD=y
CONFIG_CRYPTO_DRBG_MENU=y
CONFIG_CRYPTO_DRBG_HMAC=y
CONFIG_CRYPTO_DRBG=y
CONFIG_CRYPTO_JITTERENTROPY=y
CONFIG_CRYPTO_HASH_INFO=y
CONFIG_ASYMMETRIC_KEY_TYPE=y
CONFIG_ASYMMETRIC_PUBLIC_KEY_SUBTYPE=y
CONFIG_X509_CERTIFICATE_PARSER=y
CONFIG_PKCS8_PRIVATE_KEY_PARSER=y
CONFIG_PKCS7_MESSAGE_PARSER=y
CONFIG_PKCS7_TEST_KEY=y
CONFIG_SIGNED_PE_FILE_VERIFICATION=y
CONFIG_MODULE_SIG_KEY="certs/signing_key.pem"
CONFIG_MODULE_SIG_KEY_TYPE_RSA=y
CONFIG_SYSTEM_TRUSTED_KEYRING=y
CONFIG_SYSTEM_TRUSTED_KEYS=""
CONFIG_SYSTEM_EXTRA_CERTIFICATE=y
CONFIG_SYSTEM_EXTRA_CERTIFICATE_SIZE=4096
CONFIG_SECONDARY_TRUSTED_KEYRING=y
CONFIG_SYSTEM_BLACKLIST_KEYRING=y
CONFIG_SYSTEM_BLACKLIST_HASH_LIST=""
CONFIG_SYSTEM_REVOCATION_LIST=y
CONFIG_SYSTEM_REVOCATION_KEYS=""
CONFIG_SYSTEM_BLACKLIST_AUTH_UPDATE=y
CONFIG_BINARY_PRINTF=y
CONFIG_CRYPTO_LIB_AES=y
CONFIG_CRYPTO_LIB_SHA256=y
CONFIG_CRC_CCITT=y
CONFIG_XXHASH=y
CONFIG_AUDIT_GENERIC=y
CONFIG_LZO_COMPRESS=y
CONFIG_LZO_DECOMPRESS=y
CONFIG_ZSTD_COMMON=y
CONFIG_ZSTD_COMPRESS=y
CONFIG_ZSTD_DECOMPRESS=y
CONFIG_ASSOCIATIVE_ARRAY=y
CONFIG_SGL_ALLOC=y
CONFIG_GLOB=y
CONFIG_CLZ_TAB=y
CONFIG_MPILIB=y
CONFIG_SIGNATURE=y
CONFIG_OID_REGISTRY=y
CONFIG_STACKDEPOT=y
CONFIG_STACKDEPOT_ALWAYS_INIT=y
CONFIG_PRINTK_TIME=y
CONFIG_PRINTK_CALLER=y
CONFIG_DYNAMIC_DEBUG=y
CONFIG_DYNAMIC_DEBUG_CORE=y
CONFIG_DEBUG_INFO_DWARF5=y
CONFIG_GDB_SCRIPTS=y
CONFIG_FRAME_WARN=2048
CONFIG_READABLE_ASM=y
CONFIG_DEBUG_SECTION_MISMATCH=y
CONFIG_DEBUG_FS=y
CONFIG_DEBUG_FS_ALLOW_ALL=y
CONFIG_UBSAN=y
CONFIG_CC_HAS_UBSAN_BOUNDS=y
CONFIG_UBSAN_BOUNDS=y
CONFIG_UBSAN_ONLY_BOUNDS=y
CONFIG_UBSAN_SHIFT=y
CONFIG_UBSAN_DIV_ZERO=y
CONFIG_UBSAN_BOOL=y
CONFIG_UBSAN_ENUM=y
CONFIG_UBSAN_ALIGNMENT=y
CONFIG_PAGE_EXTENSION=y
CONFIG_DEBUG_PAGEALLOC=y
CONFIG_DEBUG_PAGEALLOC_ENABLE_DEFAULT=y
CONFIG_SLUB_DEBUG=y
CONFIG_SLUB_DEBUG_ON=y
CONFIG_PAGE_OWNER=y
CONFIG_PAGE_POISONING=y
CONFIG_DEBUG_OBJECTS=y
CONFIG_DEBUG_OBJECTS_FREE=y
CONFIG_DEBUG_OBJECTS_TIMERS=y
CONFIG_DEBUG_OBJECTS_WORK=y
CONFIG_DEBUG_OBJECTS_RCU_HEAD=y
CONFIG_DEBUG_OBJECTS_PERCPU_COUNTER=y
CONFIG_DEBUG_OBJECTS_ENABLE_DEFAULT=1
CONFIG_DEBUG_KMEMLEAK=y
CONFIG_DEBUG_KMEMLEAK_MEM_POOL_SIZE=16000
CONFIG_DEBUG_KMEMLEAK_AUTO_SCAN=y
CONFIG_DEBUG_STACK_USAGE=y
CONFIG_SCHED_STACK_END_CHECK=y
CONFIG_DEBUG_SHIRQ=y
CONFIG_PANIC_ON_OOPS=y
CONFIG_PANIC_ON_OOPS_VALUE=1
CONFIG_LOCKUP_DETECTOR=y
CONFIG_SOFTLOCKUP_DETECTOR=y
CONFIG_BOOTPARAM_SOFTLOCKUP_PANIC=y
CONFIG_DETECT_HUNG_TASK=y
CONFIG_DEFAULT_HUNG_TASK_TIMEOUT=120
CONFIG_BOOTPARAM_HUNG_TASK_PANIC=y
CONFIG_WQ_WATCHDOG=y
CONFIG_DEBUG_TIMEKEEPING=y
CONFIG_PROVE_LOCKING=y
CONFIG_PROVE_RAW_LOCK_NESTING=y
CONFIG_LOCK_STAT=y
CONFIG_DEBUG_RT_MUTEXES=y
CONFIG_DEBUG_SPINLOCK=y
CONFIG_DEBUG_MUTEXES=y
CONFIG_DEBUG_WW_MUTEX_SLOWPATH=y
CONFIG_DEBUG_RWSEMS=y
CONFIG_DEBUG_LOCK_ALLOC=y
CONFIG_LOCKDEP=y
CONFIG_LOCKDEP_BITS=15
CONFIG_LOCKDEP_CHAINS_BITS=16
CONFIG_LOCKDEP_STACK_TRACE_BITS=19
CONFIG_LOCKDEP_STACK_TRACE_HASH_BITS=14
CONFIG_LOCKDEP_CIRCULAR_QUEUE_BITS=12
CONFIG_WW_MUTEX_SELFTEST=y
CONFIG_CSD_LOCK_WAIT_DEBUG=y
CONFIG_TRACE_IRQFLAGS=y
CONFIG_DEBUG_IRQFLAGS=y
CONFIG_DEBUG_LIST=y
CONFIG_DEBUG_PLIST=y
CONFIG_DEBUG_NOTIFIERS=y
CONFIG_BUG_ON_DATA_CORRUPTION=y
CONFIG_PROVE_RCU=y
CONFIG_RCU_TRACE=y
CONFIG_NOP_TRACER=y
CONFIG_TRACE_CLOCK=y
CONFIG_RING_BUFFER=y
CONFIG_EVENT_TRACING=y
CONFIG_CONTEXT_SWITCH_TRACER=y
CONFIG_PREEMPTIRQ_TRACEPOINTS=y
CONFIG_TRACING=y
CONFIG_DRM=n
CONFIG_USB=n
CONFIG_SOUND=n
CONFIG_9P_FS=y
CONFIG_9P_FS_POSIX_ACL=y
CONFIG_9P_FS_SECURITY=y
CONFIG_ETHERNET=n
CONFIG_WLAN=n

29
kernel-configs/integrity Normal file
View File

@ -0,0 +1,29 @@
CONFIG_INTEGRITY=y
CONFIG_INTEGRITY_SIGNATURE=y
CONFIG_INTEGRITY_ASYMMETRIC_KEYS=y
CONFIG_INTEGRITY_TRUSTED_KEYRING=y
CONFIG_INTEGRITY_AUDIT=y
CONFIG_IMA=y
CONFIG_IMA_MEASURE_PCR_IDX=10
CONFIG_IMA_NG_TEMPLATE=y
CONFIG_IMA_DEFAULT_TEMPLATE="ima-ng"
CONFIG_IMA_DEFAULT_HASH_SHA256=y
CONFIG_IMA_DEFAULT_HASH="sha256"
CONFIG_IMA_WRITE_POLICY=y
CONFIG_IMA_READ_POLICY=y
CONFIG_IMA_APPRAISE=y
CONFIG_IMA_ARCH_POLICY=y
CONFIG_IMA_APPRAISE_BUILD_POLICY=y
CONFIG_IMA_APPRAISE_BOOTPARAM=y
CONFIG_IMA_APPRAISE_MODSIG=y
CONFIG_IMA_TRUSTED_KEYRING=y
CONFIG_IMA_BLACKLIST_KEYRING=y
CONFIG_IMA_LOAD_X509=y
CONFIG_IMA_X509_PATH="/etc/keys/x509_ima.der"
CONFIG_IMA_MEASURE_ASYMMETRIC_KEYS=y
CONFIG_IMA_QUEUE_EARLY_BOOT_KEYS=y
CONFIG_EVM=y
CONFIG_EVM_ATTR_FSUUID=y
CONFIG_EVM_ADD_XATTRS=y
CONFIG_EVM_LOAD_X509=y
CONFIG_EVM_X509_PATH="/etc/keys/x509_evm.der"

View File

@ -7,7 +7,7 @@ AC_DEFUN([EVMCTL_MANPAGE_DOCBOOK_XSL], [
AC_PATH_PROGS(XMLCATALOG, xmlcatalog)
AC_ARG_WITH([xml-catalog],
AC_HELP_STRING([--with-xml-catalog=CATALOG],
AS_HELP_STRING([--with-xml-catalog=CATALOG],
[path to xml catalog to use]),,
[with_xml_catalog=/etc/xml/catalog])
XML_CATALOG_FILE="$with_xml_catalog"

View File

@ -1,5 +1,5 @@
Name: ima-evm-utils
Version: 1.4
Version: 1.5
Release: 1%{?dist}
Summary: ima-evm-utils - IMA/EVM control utility
Group: System/Libraries

View File

@ -4,9 +4,17 @@ libimaevm_la_SOURCES = libimaevm.c
libimaevm_la_CPPFLAGS = $(AM_CPPFLAGS) $(LIBCRYPTO_CFLAGS)
# current[:revision[:age]]
# result: [current-age].age.revision
libimaevm_la_LDFLAGS = -version-info 3:0:0
libimaevm_la_LDFLAGS = -version-info 4:0:0
libimaevm_la_LIBADD = $(LIBCRYPTO_LIBS)
if CONFIG_SIGV1
libimaevm_la_CFLAGS = -DCONFIG_SIGV1
endif
if CONFIG_IMA_EVM_ENGINE
libimaevm_la_CFLAGS = -DCONFIG_IMA_EVM_ENGINE
endif
include_HEADERS = imaevm.h
nodist_libimaevm_la_SOURCES = hash_info.h
@ -22,6 +30,16 @@ evmctl_CPPFLAGS = $(AM_CPPFLAGS) $(LIBCRYPTO_CFLAGS)
evmctl_LDFLAGS = $(LDFLAGS_READLINE)
evmctl_LDADD = $(LIBCRYPTO_LIBS) -lkeyutils libimaevm.la
# Enable IMA signature version 1
if CONFIG_SIGV1
evmctl_CFLAGS = -DCONFIG_SIGV1
endif
# Enable "--engine" support
if CONFIG_IMA_EVM_ENGINE
evmctl_CFLAGS = -DCONFIG_IMA_EVM_ENGINE
endif
# USE_PCRTSS uses the Intel TSS
if USE_PCRTSS
evmctl_SOURCES += pcr_tss.c

File diff suppressed because it is too large Load Diff

View File

@ -48,7 +48,13 @@
#include <errno.h>
#include <sys/types.h>
#include <openssl/rsa.h>
#ifdef CONFIG_IMA_EVM_ENGINE
#include <openssl/engine.h>
#endif
#if defined(OPENSSL_NO_ENGINE) || defined(OPENSSL_NO_DYNAMIC_ENGINE)
#undef CONFIG_IMA_EVM_ENGINE
#endif
#ifdef USE_FPRINTF
#define do_log(level, fmt, args...) \
@ -85,6 +91,16 @@
#define MAX_DIGEST_SIZE 64
#define MAX_SIGNATURE_SIZE 1024
/*
* The maximum template data size is dependent on the template format. For
* example the 'ima-modsig' template includes two signatures - one for the
* entire file, the other without the appended signature - and other fields
* (e.g. file digest, file name, file digest without the appended signature).
*
* Other template formats are much smaller.
*/
#define MAX_TEMPLATE_SIZE (MAX_SIGNATURE_SIZE * 4)
#define __packed __attribute__((packed))
enum evm_ima_xattr_type {
@ -93,6 +109,7 @@ enum evm_ima_xattr_type {
EVM_IMA_XATTR_DIGSIG,
IMA_XATTR_DIGEST_NG,
EVM_XATTR_PORTABLE_DIGSIG,
IMA_VERITY_DIGSIG,
};
struct h_misc {
@ -138,7 +155,8 @@ enum digest_algo {
enum digsig_version {
DIGSIG_VERSION_1 = 1,
DIGSIG_VERSION_2
DIGSIG_VERSION_2,
DIGSIG_VERSION_3 /* hash of ima_file_id struct (portion used) */
};
struct pubkey_hdr {
@ -233,5 +251,6 @@ int ima_verify_signature(const char *file, unsigned char *sig, int siglen, unsig
void init_public_keys(const char *keyfiles);
int imaevm_hash_algo_from_sig(unsigned char *sig);
const char *imaevm_hash_algo_by_id(int algo);
int calc_hash_sigv3(enum evm_ima_xattr_type type, const char *algo, const unsigned char *in_hash, unsigned char *out_hash);
#endif

View File

@ -250,6 +250,7 @@ EVP_PKEY *read_pub_pkey(const char *keyfile, int x509)
{
FILE *fp;
EVP_PKEY *pkey = NULL;
struct stat st;
if (!keyfile)
return NULL;
@ -261,6 +262,17 @@ EVP_PKEY *read_pub_pkey(const char *keyfile, int x509)
return NULL;
}
if (fstat(fileno(fp), &st) == -1) {
log_err("Failed to fstat key file: %s\n", keyfile);
goto out;
}
if ((st.st_mode & S_IFMT) != S_IFREG) {
if (imaevm_params.verbose > LOG_INFO)
log_err("Key file is not regular file: %s\n", keyfile);
goto out;
}
if (x509) {
X509 *crt = d2i_X509_fp(fp, NULL);
@ -290,6 +302,7 @@ out:
return pkey;
}
#if CONFIG_SIGV1
RSA *read_pub_key(const char *keyfile, int x509)
{
EVP_PKEY *pkey;
@ -349,6 +362,7 @@ static int verify_hash_v1(const char *file, const unsigned char *hash, int size,
return 0;
}
#endif /* CONFIG_SIGV1 */
struct public_key_entry {
struct public_key_entry *next;
@ -422,10 +436,21 @@ void init_public_keys(const char *keyfiles)
}
/*
* Verify a signature, prefixed with the signature_v2_hdr, either based
* directly or indirectly on the file data hash.
*
* version 2: directly based on the file data hash (e.g. sha*sum)
* version 3: indirectly based on the hash of the struct ima_file_id, which
* contains the xattr type (enum evm_ima_xattr_type), the hash
* algorithm (enum hash_algo), and the file data hash
* (e.g. fsverity digest).
*
* Return: 0 verification good, 1 verification bad, -1 error.
*
* (Note: signature_v2_hdr struct does not contain the 'type'.)
*/
static int verify_hash_v2(const char *file, const unsigned char *hash, int size,
unsigned char *sig, int siglen)
static int verify_hash_common(const char *file, const unsigned char *hash,
int size, unsigned char *sig, int siglen)
{
int ret = -1;
EVP_PKEY *pkey, *pkey_free = NULL;
@ -495,6 +520,128 @@ err:
return ret;
}
/*
* Verify a signature, prefixed with the signature_v2_hdr, directly based
* on the file data hash.
*
* Return: 0 verification good, 1 verification bad, -1 error.
*/
static int verify_hash_v2(const char *file, const unsigned char *hash,
int size, unsigned char *sig, int siglen)
{
/* note: signature_v2_hdr does not contain 'type', use sig + 1 */
return verify_hash_common(file, hash, size, sig + 1, siglen - 1);
}
/*
* Verify a signature, prefixed with the signature_v2_hdr, indirectly based
* on the file data hash.
*
* Return: 0 verification good, 1 verification bad, -1 error.
*/
static int verify_hash_v3(const char *file, const unsigned char *hash,
int size, unsigned char *sig, int siglen)
{
unsigned char sigv3_hash[MAX_DIGEST_SIZE];
int ret;
ret = calc_hash_sigv3(sig[0], NULL, hash, sigv3_hash);
if (ret < 0)
return ret;
/* note: signature_v2_hdr does not contain 'type', use sig + 1 */
return verify_hash_common(file, sigv3_hash, size, sig + 1, siglen - 1);
}
#define HASH_MAX_DIGESTSIZE 64 /* kernel HASH_MAX_DIGESTSIZE is 64 bytes */
struct ima_file_id {
__u8 hash_type; /* xattr type [enum evm_ima_xattr_type] */
__u8 hash_algorithm; /* Digest algorithm [enum hash_algo] */
__u8 hash[HASH_MAX_DIGESTSIZE];
} __packed;
/*
* Calculate the signature format version 3 hash based on the portion
* of the ima_file_id structure used, not the entire structure.
*
* On success, return the hash length, otherwise for openssl errors
* return 1, other errors return -EINVAL.
*/
int calc_hash_sigv3(enum evm_ima_xattr_type type, const char *algo,
const unsigned char *in_hash, unsigned char *out_hash)
{
struct ima_file_id file_id = { .hash_type = IMA_VERITY_DIGSIG };
uint8_t *data = (uint8_t *) &file_id;
const EVP_MD *md;
EVP_MD_CTX *pctx;
unsigned int mdlen;
int err;
#if OPENSSL_VERSION_NUMBER < 0x10100000
EVP_MD_CTX ctx;
pctx = &ctx;
#else
pctx = EVP_MD_CTX_new();
#endif
int hash_algo;
int hash_size;
unsigned int unused;
if (type != IMA_VERITY_DIGSIG) {
log_err("Only fsverity supports signature format v3 (sigv3)\n");
return -EINVAL;
}
if (!algo)
algo = imaevm_params.hash_algo;
if ((hash_algo = imaevm_get_hash_algo(algo)) < 0) {
log_err("Hash algorithm %s not supported\n", algo);
return -EINVAL;
}
file_id.hash_algorithm = hash_algo;
md = EVP_get_digestbyname(algo);
if (!md) {
log_err("EVP_get_digestbyname(%s) failed\n", algo);
err = 1;
goto err;
}
hash_size = EVP_MD_size(md);
memcpy(file_id.hash, in_hash, hash_size);
err = EVP_DigestInit(pctx, md);
if (!err) {
log_err("EVP_DigestInit() failed\n");
err = 1;
goto err;
}
unused = HASH_MAX_DIGESTSIZE - hash_size;
if (!EVP_DigestUpdate(pctx, data, sizeof(file_id) - unused)) {
log_err("EVP_DigestUpdate() failed\n");
err = 1;
goto err;
}
err = EVP_DigestFinal(pctx, out_hash, &mdlen);
if (!err) {
log_err("EVP_DigestFinal() failed\n");
err = 1;
goto err;
}
err = mdlen;
err:
if (err == 1)
output_openssl_errors();
#if OPENSSL_VERSION_NUMBER >= 0x10100000
EVP_MD_CTX_free(pctx);
#endif
return err;
}
int imaevm_get_hash_algo(const char *algo)
{
int i;
@ -537,7 +684,7 @@ int imaevm_hash_algo_from_sig(unsigned char *sig)
default:
return -1;
}
} else if (sig[0] == DIGSIG_VERSION_2) {
} else if (sig[0] == DIGSIG_VERSION_2 || sig[0] == DIGSIG_VERSION_3) {
hashalgo = ((struct signature_v2_hdr *)sig)->hash_algo;
if (hashalgo >= PKEY_HASH__LAST)
return -1;
@ -546,11 +693,12 @@ int imaevm_hash_algo_from_sig(unsigned char *sig)
return -1;
}
int verify_hash(const char *file, const unsigned char *hash, int size, unsigned char *sig,
int siglen)
int verify_hash(const char *file, const unsigned char *hash, int size,
unsigned char *sig, int siglen)
{
/* Get signature type from sig header */
if (sig[0] == DIGSIG_VERSION_1) {
if (sig[1] == DIGSIG_VERSION_1) {
#if CONFIG_SIGV1
const char *key = NULL;
/* Read pubkey from RSA key */
@ -558,9 +706,16 @@ int verify_hash(const char *file, const unsigned char *hash, int size, unsigned
key = "/etc/keys/pubkey_evm.pem";
else
key = imaevm_params.keyfile;
return verify_hash_v1(file, hash, size, sig, siglen, key);
} else if (sig[0] == DIGSIG_VERSION_2) {
return verify_hash_v1(file, hash, size, sig + 1, siglen - 1,
key);
#else
log_info("Signature version 1 deprecated.");
return -1;
#endif
} else if (sig[1] == DIGSIG_VERSION_2) {
return verify_hash_v2(file, hash, size, sig, siglen);
} else if (sig[1] == DIGSIG_VERSION_3) {
return verify_hash_v3(file, hash, size, sig, siglen);
} else
return -1;
}
@ -571,11 +726,16 @@ int ima_verify_signature(const char *file, unsigned char *sig, int siglen,
unsigned char hash[MAX_DIGEST_SIZE];
int hashlen, sig_hash_algo;
if (sig[0] != EVM_IMA_XATTR_DIGSIG) {
if (sig[0] != EVM_IMA_XATTR_DIGSIG && sig[0] != IMA_VERITY_DIGSIG) {
log_err("%s: xattr ima has no signature\n", file);
return -1;
}
if (!digest && sig[0] == IMA_VERITY_DIGSIG) {
log_err("%s: calculating the fs-verity digest is not supported\n", file);
return -1;
}
sig_hash_algo = imaevm_hash_algo_from_sig(sig + 1);
if (sig_hash_algo < 0) {
log_err("%s: Invalid signature\n", file);
@ -588,17 +748,18 @@ int ima_verify_signature(const char *file, unsigned char *sig, int siglen,
* Validate the signature based on the digest included in the
* measurement list, not by calculating the local file digest.
*/
if (digestlen > 0)
return verify_hash(file, digest, digestlen, sig + 1, siglen - 1);
if (digest && digestlen > 0)
return verify_hash(file, digest, digestlen, sig, siglen);
hashlen = ima_calc_hash(file, hash);
if (hashlen <= 1)
return hashlen;
assert(hashlen <= sizeof(hash));
return verify_hash(file, hash, hashlen, sig + 1, siglen - 1);
return verify_hash(file, hash, hashlen, sig, siglen);
}
#if CONFIG_SIGV1
/*
* Create binary key representation suitable for kernel
*/
@ -657,6 +818,7 @@ void calc_keyid_v1(uint8_t *keyid, char *str, const unsigned char *pkey, int len
if (imaevm_params.verbose > LOG_INFO)
log_info("keyid-v1: %s\n", str);
}
#endif /* CONFIG_SIGV1 */
/*
* Calculate keyid of the public_key part of EVP_PKEY
@ -761,6 +923,7 @@ static int read_keyid_from_cert(uint32_t *keyid_be, const char *certfile, int tr
ERR_print_errors_fp(stderr);
log_err("read keyid: %s: Error reading x509 certificate\n",
certfile);
return -1;
}
if (!(skid = x509_get_skid(x, &skid_len))) {
@ -803,9 +966,10 @@ uint32_t imaevm_read_keyid(const char *certfile)
static EVP_PKEY *read_priv_pkey(const char *keyfile, const char *keypass)
{
FILE *fp;
EVP_PKEY *pkey;
EVP_PKEY *pkey = NULL;
if (!strncmp(keyfile, "pkcs11:", 7)) {
#ifdef CONFIG_IMA_EVM_ENGINE
if (!imaevm_params.keyid) {
log_err("When using a pkcs11 URI you must provide the keyid with an option\n");
return NULL;
@ -822,6 +986,10 @@ static EVP_PKEY *read_priv_pkey(const char *keyfile, const char *keypass)
log_err("Failed to load private key %s\n", keyfile);
goto err_engine;
}
#else
log_err("OpenSSL \"engine\" support is disabled\n");
goto err_engine;
#endif
} else {
fp = fopen(keyfile, "r");
if (!fp) {
@ -845,6 +1013,7 @@ err_engine:
return NULL;
}
#if CONFIG_SIGV1
static RSA *read_priv_key(const char *keyfile, const char *keypass)
{
EVP_PKEY *pkey;
@ -955,6 +1124,7 @@ out:
RSA_free(key);
return len;
}
#endif /* CONFIG_SIGV1 */
/*
* @sig is assumed to be of (MAX_SIGNATURE_SIZE - 1) size
@ -1069,9 +1239,14 @@ int sign_hash(const char *hashalgo, const unsigned char *hash, int size, const c
if (keypass)
imaevm_params.keypass = keypass;
return imaevm_params.x509 ?
sign_hash_v2(hashalgo, hash, size, keyfile, sig) :
sign_hash_v1(hashalgo, hash, size, keyfile, sig);
if (imaevm_params.x509)
return sign_hash_v2(hashalgo, hash, size, keyfile, sig);
#if CONFIG_SIGV1
else
return sign_hash_v1(hashalgo, hash, size, keyfile, sig);
#endif
log_info("Signature version 1 deprecated.");
return -1;
}
static void libinit()

View File

@ -20,21 +20,11 @@
#undef MAX_DIGEST_SIZE /* imaevm uses a different value than the TSS */
#include <ibmtss/tss.h>
#define CMD "tsspcrread"
static char path[PATH_MAX];
int tpm2_pcr_supported(void)
{
if (imaevm_params.verbose > LOG_INFO)
log_info("Using %s to read PCRs.\n", CMD);
log_info("Using ibmtss to read PCRs\n");
if (get_cmd_path(CMD, path, sizeof(path))) {
log_debug("Couldn't find '%s' in $PATH\n", CMD);
return 0;
}
log_debug("Found '%s' in $PATH\n", CMD);
return 1;
}

View File

@ -60,11 +60,11 @@ int tpm2_pcr_supported(void)
log_info("Using %s to read PCRs.\n", CMD);
if (get_cmd_path(CMD, path, sizeof(path))) {
log_debug("Couldn't find '%s' in $PATH\n", CMD);
log_info("Couldn't find '%s' in %s\n", CMD, path);
return 0;
}
log_debug("Found '%s' in $PATH\n", CMD);
log_debug("Found '%s' in %s\n", CMD, path);
return 1;
}

View File

@ -1,7 +1,24 @@
check_SCRIPTS =
TESTS = $(check_SCRIPTS)
check_SCRIPTS += ima_hash.test sign_verify.test boot_aggregate.test
check_SCRIPTS += ima_hash.test sign_verify.test boot_aggregate.test \
fsverity.test portable_signatures.test ima_policy_check.test \
mmap_check.test
check_PROGRAMS := test_mmap
.PHONY: check_logs
check_logs:
@for log in $(TEST_LOGS); do \
echo -e "\n***" $$log "***" ; \
case $$log in \
ima_hash.log | sign_verify.log ) \
tail -3 $$log ; \
grep "skipped" $$log && grep "skipped" $$log | wc -l ;; \
*) \
cat $$log ;; \
esac ; \
done
clean-local:
-rm -f *.txt *.out *.sig *.sig2

View File

@ -12,7 +12,7 @@
# for verifying the calculated boot_aggregate is included in this
# directory as well.
trap cleanup SIGINT SIGTERM EXIT
trap '_report_exit_and_cleanup cleanup' SIGINT SIGTERM EXIT
# Base VERBOSE on the environment variable, if set.
VERBOSE="${VERBOSE:-0}"
@ -126,8 +126,10 @@ display_pcrs() {
# Verify that the last "boot_aggregate" record in the IMA measurement
# list matches.
check() {
local options=$1
echo "INFO: Calculating the boot_aggregate (PCRs 0 - 9) for multiple banks"
bootaggr=$(evmctl ima_boot_aggregate)
bootaggr=$(evmctl ima_boot_aggregate ${options})
if [ $? -ne 0 ]; then
echo "${CYAN}SKIP: evmctl ima_boot_aggregate: $bootaggr${NORM}"
exit "$SKIP"
@ -151,6 +153,7 @@ check() {
}
if [ "$(id -u)" = 0 ] && [ -c "/dev/tpm0" ]; then
BOOTAGGR_OPTIONS="--hwtpm"
ASCII_RUNTIME_MEASUREMENTS="/sys/kernel/security/ima/ascii_runtime_measurements"
if [ ! -d "/sys/kernel/security/ima" ]; then
echo "${CYAN}SKIP: CONFIG_IMA not enabled${NORM}"
@ -194,4 +197,4 @@ if [ "$(id -u)" != 0 ] || [ ! -c "/dev/tpm0" ]; then
fi
fi
expect_pass check
expect_pass check $BOOTAGGR_OPTIONS

385
tests/fsverity.test Executable file
View File

@ -0,0 +1,385 @@
#!/bin/bash
# SPDX-License-Identifier: GPL-2.0
#
# Test IMA support for including fs-verity enabled files measurements
# in the IMA measurement list.
#
# Define policy rules showing the different types of IMA and fs-verity
# records in the IMA measurement list. Include examples of files that
# are suppose to be fs-verity enabled, but aren't.
#
# test 1: IMA policy rule using the new ima-ngv2 template
# - Hash prefixed with "ima:"
#
# test 2: fs-verity IMA policy rule using the new ima-ngv2 template
# - fs-verity hash prefixed with "verity:"
# - Non fs-verity enabled file, zeros prefixed with "verity:"
#
# test 3: IMA policy rule using the new ima-sigv2 template
# - Hash prefixed with "ima:"
# - Appended signature, when available.
#
# test 4: fs-verity IMA policy rule using the new ima-sigv2 template
# - fs-verity hash prefixed with "verity:"
# - Non fs-verity enabled file, zeros prefixed with "verity:"
# - Appended IMA signature of fs-verity file hash, when available.
# To avoid affecting the system's IMA custom policy or requiring a
# reboot between tests, define policy rules based on UUID. However,
# since the policy rules are walked sequentially, the system's IMA
# custom policy rules might take precedence.
cd "$(dirname "$0")" || exit 1
PATH=../src:../fsverity-utils:$PATH
source ./functions.sh
# Base VERBOSE on the environment variable, if set.
VERBOSE="${VERBOSE:-0}"
IMA_POLICY_FILE="/sys/kernel/security/integrity/ima/policy"
IMA_MEASUREMENT_LIST="/sys/kernel/security/integrity/ima/ascii_runtime_measurements"
TST_MNT="/tmp/fsverity-test"
TST_IMG="/tmp/test.img"
LOOPBACK_MOUNTED=0
FSVERITY="$(which fsverity)"
_require dd mkfs blkid e2fsck tune2fs evmctl setfattr
./gen-keys.sh >/dev/null 2>&1
trap '_report_exit_and_cleanup _cleanup_env cleanup' SIGINT SIGTERM EXIT
cleanup() {
if [ -e $TST_MNT ]; then
if [ $LOOPBACK_MOUNTED -eq 1 ]; then
umount $TST_MNT
fi
if [ -f "$TST_IMG" ]; then
rm "$TST_IMG"
fi
fi
}
# Loopback mount a file
mount_loopback_file() {
local ret
if [ ! -d $TST_MNT ]; then
mkdir $TST_MNT
fi
# if modprobe loop; then
# echo "${CYAN}INFO: modprobe loop failed${NORM}"
# fi
if ! losetup -f &> /dev/null; then
echo "${RED}FAILURE: losetup${NORM}"
exit "$FAIL"
fi
mount -v -o loop ${TST_IMG} $TST_MNT
ret=$?
if [ "${ret}" -eq 0 ]; then
LOOPBACK_MOUNTED=1
fi
return "$ret"
}
# Change the loopback mounted filesystem's UUID in between tests
change_loopback_file_uuid() {
echo " "
[ "$VERBOSE" -ge 1 ] && echo "INFO: Changing loopback file uuid"
umount $TST_MNT
if ! e2fsck -y -f ${TST_IMG} &> /dev/null; then
echo "${RED}FAILURE: e2fsck${NORM}"
exit "$FAIL"
fi
if ! tune2fs -f ${TST_IMG} -U random &> /dev/null; then
echo "${RED}FAILURE: change UUID${NORM}"
exit "$FAIL"
fi
[ "$VERBOSE" -ge 1 ] && echo "INFO: Remounting loopback filesystem"
if ! mount_loopback_file; then
echo "${RED}FAILURE: re-mounting loopback filesystem${NORM}"
exit "$FAIL"
fi
return 0
}
# Create a file to be loopback mounted
create_loopback_file() {
local fs_type=$1
local options=""
echo "INFO: Creating loopback filesystem"
case $fs_type in
ext4|f2fs)
options="-O verity"
;;
btrfs)
;;
*)
echo "${RED}FAILURE: unsupported fs-verity filesystem${NORM}"
exit "${FAIL}"
;;
esac
[ "$VERBOSE" -ge 2 ] && echo "INFO: Creating a file to be loopback mounted with options: $options"
if ! dd if=/dev/zero of="${TST_IMG}" bs=100M count=6 &> /dev/null; then
echo "${RED}FAILURE: creating ${TST_IMG}${NORM}"
exit "$FAIL"
fi
echo "INFO: Building an $fs_type filesystem"
if ! mkfs -t "$fs_type" -q "${TST_IMG}" "$options"; then
echo "${RED}FAILURE: Creating $fs_type filesystem${NORM}"
exit "$FAIL"
fi
echo "INFO: Mounting loopback filesystem"
if ! mount_loopback_file; then
echo "${RED}FAILURE: mounting loopback filesystem${NORM}"
exit "$FAIL"
fi
return 0
}
get_current_uuid() {
[ "$VERBOSE" -ge 2 ] && echo "INFO: Getting loopback file uuid"
if ! UUID=$(blkid -s UUID -o value ${TST_IMG}); then
echo "${RED}FAILURE: to get UUID${NORM}"
return "$FAIL"
fi
return 0
}
unqualified_bprm_rule() {
local test=$1
local rule=$2
local rule_match="measure func=BPRM_CHECK"
local rule_dontmatch="fsuuid"
if [ -z "${rule##*$digest_type=verity*}" ]; then
if grep "$rule_match" $IMA_POLICY_FILE | grep -v "$rule_dontmatch" &> /dev/null; then
return "$SKIP"
fi
fi
return 0
}
load_policy_rule() {
local test=$1
local rule=$2
if ! get_current_uuid; then
echo "${RED}FAILURE:FAILED getting uuid${NORM}"
exit "$FAIL"
fi
unqualified_bprm_rule "${test}" "${rule}"
if [ $? -eq "${SKIP}" ]; then
echo "${CYAN}SKIP: fsuuid unqualified \"BPRM_CHECK\" rule exists${NORM}"
return "$SKIP"
fi
echo "$test: rule: $rule fsuuid=$UUID"
if ! echo "$rule fsuuid=$UUID" > $IMA_POLICY_FILE; then
echo "${CYAN}SKIP: Loading policy rule failed, skipping test${NORM}"
return "$SKIP"
fi
return 0
}
create_file() {
local test=$1
local type=$2
TST_FILE=$(mktemp -p $TST_MNT -t "${type}".XXXXXX)
[ "$VERBOSE" -ge 1 ] && echo "INFO: creating $TST_FILE"
# heredoc to create a script
cat <<-EOF > "$TST_FILE"
#!/bin/bash
echo "Hello" &> /dev/null
EOF
chmod a+x "$TST_FILE"
}
measure-verity() {
local test=$1
local verity="${2:-disabled}"
local digest_filename
local error="$OK"
local KEY=$PWD/test-rsa2048.key
create_file "$test" verity-hash
if [ "$verity" = "enabled" ]; then
msg="Measuring fs-verity enabled file $TST_FILE"
if ! "$FSVERITY" enable "$TST_FILE" &> /dev/null; then
echo "${CYAN}SKIP: Failed enabling fs-verity on $TST_FILE${NORM}"
return "$SKIP"
fi
else
msg="Measuring non fs-verity enabled file $TST_FILE"
fi
# Sign the fsverity digest and write it as security.ima xattr.
# "evmctl sign_hash" input: <digest> <filename>
# "evmctl sign_hash" output: <digest> <filename> <signature>
[ "$VERBOSE" -ge 2 ] && echo "INFO: Signing the fsverity digest"
xattr=$("$FSVERITY" digest "$TST_FILE" | evmctl sign_hash --veritysig --key "$KEY" 2> /dev/null)
sig=$(echo "$xattr" | cut -d' ' -f3)
# On failure to write security.ima xattr, the signature will simply
# not be appended to the measurement list record.
if ! setfattr -n security.ima -v "0x$sig" "$TST_FILE"; then
echo "${CYAN}INFO: Failed to write security.ima xattr${NORM}"
fi
"$TST_FILE"
# "fsverity digest" calculates the fsverity hash, even for
# non fs-verity enabled files.
digest_filename=$("$FSVERITY" digest "$TST_FILE")
[ "$VERBOSE" -ge 2 ] && echo "INFO: verity:$digest_filename"
grep "verity:$digest_filename" $IMA_MEASUREMENT_LIST &> /dev/null
ret=$?
# Not finding the "fsverity digest" result in the IMA measurement
# list is expected for non fs-verity enabled files. The measurement
# list will contain zeros for the file hash.
if [ $ret -eq 1 ]; then
error="$FAIL"
if [ "$verity" = "enabled" ]; then
echo "${RED}FAILURE: ${msg} ${NORM}"
else
echo "${GREEN}SUCCESS: ${msg}, fsverity digest not found${NORM}"
fi
else
if [ "$verity" = "enabled" ]; then
echo "${GREEN}SUCCESS: ${msg} ${NORM}"
else
error="$FAIL"
echo "${RED}FAILURE: ${msg} ${NORM}"
fi
fi
return "$error"
}
measure-ima() {
local test=$1
local digest_filename
local error="$OK"
local hashalg
local digestsum
create_file "$test" ima-hash
"$TST_FILE"
hashalg=$(grep "${TST_FILE}" $IMA_MEASUREMENT_LIST | cut -d':' -f2)
if [ -z "${hashalg}" ]; then
echo "${CYAN}SKIP: Measurement record with algorithm not found${NORM}"
return "$SKIP"
fi
digestsum=$(which "${hashalg}"sum)
if [ -z "${digestsum}" ]; then
echo "${CYAN}SKIP: ${hashalg}sum is not installed${NORM}"
return "$SKIP"
fi
# sha1sum,sha256sum return: <digest> <2 spaces> <filename>
# Remove the extra space before the filename
digest_filename=$(${digestsum} "$TST_FILE" | sed "s/\ \ /\ /")
[ "$VERBOSE" -ge 2 ] && echo "$test: $digest_filename"
if grep "$digest_filename" $IMA_MEASUREMENT_LIST &> /dev/null; then
echo "${GREEN}SUCCESS: Measuring $TST_FILE ${NORM}"
else
error="$FAIL"
echo "${RED}FAILURE: Measuring $TST_FILE ${NORM}"
fi
return "$error"
}
# Run in the new environment if TST_ENV is set.
_run_env "$TST_KERNEL" "$PWD/$(basename "$0")" "TST_ENV=$TST_ENV TST_KERNEL=$TST_KERNEL PATH=$PATH LD_LIBRARY_PATH=$LD_LIBRARY_PATH VERBOSE=$VERBOSE"
# Exit from the creator of the new environment.
_exit_env "$TST_KERNEL"
# Mount filesystems in the new environment.
_init_env
# Dependency on being able to read and write the IMA policy file.
# Requires both CONFIG_IMA_WRITE_POLICY, CONFIG_IMA_READ_POLICY be
# enabled.
if [ -e "$IMA_POLICY_FILE" ]; then
mode=$(stat -c "%a" $IMA_POLICY_FILE)
if [ "$mode" != "600" ]; then
echo "${CYAN}SKIP: IMA policy file must be read-write${NORM}"
exit "$SKIP"
fi
else
echo "${CYAN}SKIP: $IMA_POLICY_FILE does not exist${NORM}"
exit "$SKIP"
fi
# Skip the test if fsverity is not found; using _require fails the test.
if [ -z "$FSVERITY" ]; then
echo "${CYAN}SKIP: fsverity is not installed${NORM}"
exit "$SKIP"
fi
if [ "x$(id -u)" != "x0" ]; then
echo "${CYAN}SKIP: Must be root to execute this test${NORM}"
exit "$SKIP"
fi
create_loopback_file ext4
# Commit 989dc72511f7 ("ima: define a new template field named 'd-ngv2' and
# templates") introduced ima-ngv2 and ima-sigv2 in linux-5.19.
__skip() { return "$SKIP"; }
# IMA policy rule using the ima-ngv2 template
if load_policy_rule test1 "measure func=BPRM_CHECK template=ima-ngv2"; then
expect_pass measure-ima test1
else
expect_pass __skip
fi
# fsverity IMA policy rule using the ima-ngv2 template
change_loopback_file_uuid
if load_policy_rule test2 "measure func=BPRM_CHECK template=ima-ngv2 digest_type=verity"; then
expect_fail measure-verity test2
expect_pass measure-verity test2 enabled
else
expect_pass __skip
expect_pass __skip
fi
# IMA policy rule using the ima-sigv2 template
change_loopback_file_uuid
if load_policy_rule test3 "measure func=BPRM_CHECK template=ima-sigv2"; then
expect_pass measure-ima test3
else
expect_pass __skip
fi
# fsverity IMA policy rule using the ima-sigv2 template
change_loopback_file_uuid
if load_policy_rule test4 "measure func=BPRM_CHECK template=ima-sigv2 digest_type=verity"; then
expect_fail measure-verity test4
expect_pass measure-verity test4 enabled
else
expect_pass __skip
expect_pass __skip
fi
exit

View File

@ -72,6 +72,12 @@ declare -i TNESTED=0 # just for sanity checking
expect_pass() {
local -i ret
if [ -n "$TST_LIST" ] && [ "${TST_LIST/$1/}" = "$TST_LIST" ]; then
[ "$VERBOSE" -gt 1 ] && echo "____ SKIP test: $*"
testsskip+=1
return "$SKIP"
fi
if [ $TNESTED -gt 0 ]; then
echo $RED"expect_pass should not be run nested"$NORM
testsfail+=1
@ -94,10 +100,35 @@ expect_pass() {
return $ret
}
expect_pass_if() {
local indexes="$1"
local ret idx
shift
expect_pass "$@"
ret=$?
if [ $ret -ne 0 ] && [ $ret -ne 77 ] && [ -n "$PATCHES" ]; then
echo $YELLOW"Possibly missing patches:"$NORM
for idx in $indexes; do
echo $YELLOW" - ${PATCHES[$((idx))]}"$NORM
done
fi
return $ret
}
# Eval negative test (one that should fail) and account its result
expect_fail() {
local ret
if [ -n "$TST_LIST" ] && [ "${TST_LIST/$1/}" = "$TST_LIST" ]; then
[ "$VERBOSE" -gt 1 ] && echo "____ SKIP test: $*"
testsskip+=1
return "$SKIP"
fi
if [ $TNESTED -gt 0 ]; then
echo $RED"expect_fail should not be run nested"$NORM
testsfail+=1
@ -125,6 +156,25 @@ expect_fail() {
return $ret
}
expect_fail_if() {
local indexes="$1"
local ret idx
shift
expect_fail "$@"
ret=$?
if { [ $ret -eq 0 ] || [ $ret -eq 99 ]; } && [ -n "$PATCHES" ]; then
echo $YELLOW"Possibly missing patches:"$NORM
for idx in $indexes; do
echo $YELLOW" - ${PATCHES[$((idx))]}"$NORM
done
fi
return $ret
}
# return true if current test is positive
_test_expected_to_pass() {
[ ! $TFAIL ]
@ -250,10 +300,14 @@ _enable_gost_engine() {
# Show test stats and exit into automake test system
# with proper exit code (same as ours). Do cleanups.
_report_exit_and_cleanup() {
local exit_code=$?
if [ -n "${WORKDIR}" ]; then
rm -rf "${WORKDIR}"
fi
"$@"
if [ $testsfail -gt 0 ]; then
echo "================================="
echo " Run with FAILEARLY=1 $0 $*"
@ -267,12 +321,33 @@ _report_exit_and_cleanup() {
[ $testsfail -gt 0 ] && echo -n "$RED" || echo -n "$NORM"
echo " FAIL: $testsfail"
echo "$NORM"
# Signal failure to the testing environment creator with an unclean shutdown.
if [ -n "$TST_ENV" ] && [ $$ -eq 1 ]; then
if [ -z "$(command -v poweroff)" ]; then
echo "Warning: cannot properly shutdown system"
fi
# If no test was executed and the script was successful,
# do a clean shutdown.
if [ $testsfail -eq 0 ] && [ $testspass -eq 0 ] && [ $testsskip -eq 0 ] &&
[ $exit_code -ne "$FAIL" ] && [ $exit_code -ne "$HARDFAIL" ]; then
poweroff -f
fi
# If tests were executed and no test failed, do a clean shutdown.
if { [ $testspass -gt 0 ] || [ $testsskip -gt 0 ]; } &&
[ $testsfail -eq 0 ]; then
poweroff -f
fi
fi
if [ $testsfail -gt 0 ]; then
exit "$FAIL"
elif [ $testspass -gt 0 ]; then
exit "$OK"
else
elif [ $testsskip -gt 0 ]; then
exit "$SKIP"
else
exit "$exit_code"
fi
}
@ -313,3 +388,75 @@ _softhsm_teardown() {
unset SOFTHSM_SETUP_CONFIGDIR SOFTHSM2_CONF PKCS11_KEYURI \
EVMCTL_ENGINE OPENSSL_ENGINE OPENSSL_KEYFORM
}
# Syntax: _run_env <kernel> <init> <additional kernel parameters>
_run_env() {
if [ -z "$TST_ENV" ]; then
return
fi
if [ $$ -eq 1 ]; then
return
fi
if [ "$TST_ENV" = "um" ]; then
expect_pass "$1" rootfstype=hostfs rw init="$2" quiet mem=2048M "$3"
else
echo $RED"Testing environment $TST_ENV not supported"$NORM
exit "$FAIL"
fi
}
# Syntax: _exit_env <kernel>
_exit_env() {
if [ -z "$TST_ENV" ]; then
return
fi
if [ $$ -eq 1 ]; then
return
fi
exit "$OK"
}
# Syntax: _init_env
_init_env() {
if [ -z "$TST_ENV" ]; then
return
fi
if [ $$ -ne 1 ]; then
return
fi
mount -t tmpfs tmpfs /tmp
mount -t proc proc /proc
mount -t sysfs sysfs /sys
mount -t securityfs securityfs /sys/kernel/security
if [ -n "$(command -v haveged 2> /dev/null)" ]; then
$(command -v haveged) -w 1024 &> /dev/null
fi
pushd "$PWD" > /dev/null || exit "$FAIL"
}
# Syntax: _cleanup_env <cleanup function>
_cleanup_env() {
if [ -z "$TST_ENV" ]; then
$1
return
fi
if [ $$ -ne 1 ]; then
return
fi
$1
umount /sys/kernel/security
umount /sys
umount /proc
umount /tmp
}

View File

@ -71,7 +71,7 @@ for m in 1024 1024_skid 2048; do
ext=
fi
if [ ! -e test-rsa$m.key ]; then
log openssl req -verbose -new -nodes -utf8 -sha1 -days 10000 -batch -x509 $ext \
log openssl req -verbose -new -nodes -utf8 -sha256 -days 10000 -batch -x509 $ext \
-config test-ca.conf \
-newkey rsa:$bits \
-out test-rsa$m.cer -outform DER \
@ -93,7 +93,7 @@ for curve in prime192v1 prime256v1; do
continue
fi
if [ ! -e test-$curve.key ]; then
log openssl req -verbose -new -nodes -utf8 -sha1 -days 10000 -batch -x509 \
log openssl req -verbose -new -nodes -utf8 -sha256 -days 10000 -batch -x509 \
-config test-ca.conf \
-newkey ec \
-pkeyopt ec_paramgen_curve:$curve \

211
tests/ima_policy_check.awk Executable file
View File

@ -0,0 +1,211 @@
#! /usr/bin/gawk -f
# SPDX-License-Identifier: GPL-2.0
#
# Copyright (C) 2023 Roberto Sassu <roberto.sassu@huawei.com>
#
# Check a new rule against the loaded IMA policy.
#
# Documentation/ABI/testing/ima_policy (Linux kernel)
# base: [[func=] [mask=] [fsmagic=] [fsuuid=] [fsname=]
# [uid=] [euid=] [gid=] [egid=]
# [fowner=] [fgroup=]]
# lsm: [[subj_user=] [subj_role=] [subj_type=]
# [obj_user=] [obj_role=] [obj_type=]]
# option: [digest_type=] [template=] [permit_directio]
# [appraise_type=] [appraise_flag=]
# [appraise_algos=] [keyrings=]
#
# Rules don't overlap if their actions are unrelated (cannot be matched without
# dont_) and there is no combination of appraise with another do action (e.g.
# measure, audit, hash). The second condition is due to the fact that appraise
# might still forbid other actions expected to be performed by a test that did
# not setup appraisal. Checking appraise for new rules is not sufficient,
# because that rule could be added anyway. By checking existing rules as well,
# a warning will be displayed when tests inserting rules with other do actions
# are reexecuted.
#
# Also, rules don't overlap if both include the same policy keyword(s) (in base
# or lsm, except func), at least one, with a different value. Different func
# values don't imply non-overlap, due to the fact that a test command might
# cause the execution of multiple hooks (e.g. FILE_CHECK in addition to
# MMAP_CHECK). Despite one test is willing to test a particular hook, this could
# have side effects on other tests (e.g. one test sets: appraise func=MMAP_CHECK
# and another: measure func=FILE_CHECK; the second test might see an unexpected
# measurement due to the first test being executed; or the second test cannot
# unexpectedly do mmap).
#
# Currently, the < > operators are not supported and overlapping is asserted
# even if intervals are disjoint. If supported, non-overlapping conditions could
# be found. With the ^ modifier, no disjoint intervals can be found. Overlapping
# is always reported.
#
# Rule equivalence is determined by checking each key/value pair, regardless of
# their order. However, the action must always be at the beginning of the rules.
# Rules with aliases are considered equivalent to those with their source (e.g.
# rules with PATH_CHECK and FILE_MMAP are considered as equivalent to rules with
# FILE_CHECK and MMAP_CHECK).
#
# Return a bit mask with the following values:
# - 1: invalid new rule;
# - 2: overlap of the new rule with an existing rule in the IMA policy;
# - 4: new rule exists in the IMA policy.
BEGIN {
# Policy definitions.
actions_str="measure dont_measure appraise dont_appraise audit hash dont_hash"
split(actions_str, actions_array);
keywords_str="func mask fsmagic fsuuid fsname uid euid gid egid fowner fgroup subj_user subj_role subj_type obj_user obj_role obj_type";
split(keywords_str, keywords_array);
options_str="digest_type template permit_directio appraise_type appraise_flag appraise_algos keyrings";
split(options_str, options_array);
# Key types.
key_type_unknown=0;
key_type_action=1;
key_type_keyword=2;
key_type_option=3;
# Result values.
ret_invalid_rule=1;
ret_rule_overlap=2;
ret_same_rule_exists=4;
for (action_idx in actions_array)
key_types[actions_array[action_idx]]=key_type_action;
for (keyword_idx in keywords_array)
key_types[keywords_array[keyword_idx]]=key_type_keyword;
for (option_idx in options_array)
key_types[options_array[option_idx]]=key_type_option;
new_rule=1;
result=0;
}
{
# Delete arrays from previous rule.
if (!new_rule) {
delete current_rule_array;
delete current_rule_operator_array;
}
# Check empty rules.
if (!length($0)) {
if (new_rule) {
result=or(result, ret_invalid_rule);
exit;
}
next;
}
for (i=1; i<=NF; i++) {
# Parse key/value pair.
split($i, key_value_array, /[=,>,<]/, separator_array);
key=key_value_array[1];
value=key_value_array[2];
if (key == "func") {
# Normalize values of IMA hooks to what IMA will print.
if (value == "FILE_MMAP")
value="MMAP_CHECK";
else if (value == "PATH_CHECK")
value="FILE_CHECK";
}
# Basic validity check (not necessary in general for the IMA policy, but useful to find typos in the tests).
if (key_types[key] == key_type_unknown ||
(i == 1 && key_types[key] != key_type_action)) {
result=or(result, ret_invalid_rule);
exit;
}
# Store key/value pair and operator into an array.
if (new_rule) {
new_rule_array[key]=value;
new_rule_operator_array[key]=separator_array[1];
} else {
current_rule_array[key]=value;
current_rule_operator_array[key]=separator_array[1];
}
# Store original action and action without dont_.
if (i == 1) {
if (new_rule) {
new_rule_action=key;
new_rule_action_sub=key;
gsub(/dont_/, "", new_rule_action_sub);
} else {
current_rule_action=key;
current_rule_action_sub=key;
gsub(/dont_/, "", current_rule_action_sub);
}
}
}
# Go to the next line, to compare the new rule with rules in the IMA policy.
if (new_rule) {
new_rule=0;
next;
}
# No overlap by action (unrelated rules and no combination appraise - <do action>), new rule safe to add to the IMA policy.
if (current_rule_action_sub != new_rule_action_sub &&
(current_rule_action != "appraise" || new_rule_action ~ /^dont_/) &&
(new_rule_action != "appraise" || current_rule_action ~ /^dont_/))
next;
same_rule=1;
overlap_rule=1;
for (key in key_types) {
if (!(key in new_rule_array)) {
# Key in current rule but not in new rule.
if (key in current_rule_array)
same_rule=0;
# Key not in new rule and not in current rule.
continue;
}
if (!(key in current_rule_array)) {
# Key in new rule but not in current rule.
if (key in new_rule_array)
same_rule=0;
# Key not in current rule and not in new rule.
continue;
}
# Same value and operator.
if (new_rule_array[key] == current_rule_array[key] &&
new_rule_operator_array[key] == current_rule_operator_array[key])
continue;
# Different value and/or operator.
same_rule=0;
# Not a policy keyword, not useful to determine overlap.
if (key_types[key] != key_type_keyword)
continue;
# > < operators are not supported, cannot determine overlap.
if (new_rule_operator_array[key] != "=" || current_rule_operator_array[key] != "=")
continue;
# ^ modifier does not make disjoint sets, cannot determine overlap.
if (new_rule_array[key] ~ /^\^/ || current_rule_array[key] ~ /^\^/)
continue;
# One test command can invoke multiple hooks, cannot determine overlap from func.
if (key == "func")
continue;
# No overlap by policy keyword, new rule safe to add to the IMA policy.
overlap_rule=0;
next;
}
if (same_rule)
result=or(result, ret_same_rule_exists);
else if (overlap_rule)
result=or(result, ret_rule_overlap);
}
END {
exit result;
}

245
tests/ima_policy_check.test Executable file
View File

@ -0,0 +1,245 @@
#!/bin/bash
# SPDX-License-Identifier: GPL-2.0
#
# Copyright (C) 2023 Roberto Sassu <roberto.sassu@huawei.com>
#
# Test for ima_policy_check.awk
trap '_report_exit_and_cleanup' SIGINT SIGTERM EXIT
cd "$(dirname "$0")" || exit 1
. ./functions.sh
export PATH=$PWD:$PATH
check_result() {
local result
echo -e "\nTest: $1"
echo "New rule: $2"
echo "IMA policy: $3"
echo -n "Result (expect $4): "
echo -e "$2\n$3" | ima_policy_check.awk
result=$?
if [ "$result" -ne "$4" ]; then
echo "${RED}$result${NORM}"
return "$FAIL"
fi
echo "${GREEN}$result${NORM}"
return "$OK"
}
# ima_policy_check.awk returns a bit mask with the following values:
# - 1: invalid new rule;
# - 2: overlap of the new rule with an existing rule in the IMA policy;
# - 4: new rule exists in the IMA policy.
# Basic checks.
desc="empty IMA policy"
rule="measure func=FILE_CHECK"
ima_policy=""
expect_pass check_result "$desc" "$rule" "$ima_policy" 0
desc="Empty new rule"
rule=""
ima_policy=""
expect_pass check_result "$desc" "$rule" "$ima_policy" 1
desc="Unknown policy keyword fun"
rule="measure fun=FILE_CHECK"
ima_policy=""
expect_pass check_result "$desc" "$rule" "$ima_policy" 1
desc="Missing action"
rule="func=FILE_CHECK"
ima_policy=""
expect_pass check_result "$desc" "$rule" "$ima_policy" 1
# Non-overlapping rules.
desc="Non-overlapping by action measure/dont_appraise, same func"
rule="measure func=FILE_CHECK"
ima_policy="dont_appraise func=FILE_CHECK"
expect_pass check_result "$desc" "$rule" "$ima_policy" 0
desc="Non-overlapping by action audit/dont_appraise, same func"
rule="audit func=FILE_CHECK"
ima_policy="dont_appraise func=FILE_CHECK"
expect_pass check_result "$desc" "$rule" "$ima_policy" 0
desc="Non-overlapping by action appraise/dont_measure, same func"
rule="appraise func=FILE_CHECK"
ima_policy="dont_measure func=FILE_CHECK"
expect_pass check_result "$desc" "$rule" "$ima_policy" 0
desc="Non-overlapping by action dont_measure/hash, same func"
rule="dont_measure func=FILE_CHECK"
ima_policy="hash func=FILE_CHECK"
expect_pass check_result "$desc" "$rule" "$ima_policy" 0
desc="Non-overlapping by uid, func is equal"
rule="measure func=FILE_CHECK uid=0"
ima_policy="measure uid=1 func=FILE_CHECK"
expect_pass check_result "$desc" "$rule" "$ima_policy" 0
desc="Non-overlapping by uid, func is equal, same policy options"
rule="measure func=FILE_CHECK uid=0 permit_directio"
ima_policy="measure uid=1 func=FILE_CHECK permit_directio"
expect_pass check_result "$desc" "$rule" "$ima_policy" 0
desc="Non-overlapping by mask, func and uid are equal, same policy options"
rule="measure func=FILE_CHECK uid=0 permit_directio mask=MAY_READ"
ima_policy="measure uid=0 mask=MAY_EXEC func=FILE_CHECK permit_directio"
expect_pass check_result "$desc" "$rule" "$ima_policy" 0
desc="Non-overlapping by mask, func and uid are equal, different policy options"
rule="measure func=FILE_CHECK uid=0 permit_directio mask=MAY_READ"
ima_policy="measure uid=0 mask=MAY_EXEC func=FILE_CHECK"
expect_pass check_result "$desc" "$rule" "$ima_policy" 0
# Overlapping and different rules.
desc="same actions, different keywords"
rule="appraise func=FILE_CHECK"
ima_policy="appraise uid=0"
expect_pass check_result "$desc" "$rule" "$ima_policy" 2
desc="unrelated actions with appraise and a do action, same func"
rule="appraise func=FILE_CHECK"
ima_policy="measure func=FILE_CHECK"
expect_pass check_result "$desc" "$rule" "$ima_policy" 2
desc="unrelated actions with appraise and a do action, different func"
rule="appraise func=FILE_CHECK"
ima_policy="measure func=MMAP_CHECK"
expect_pass check_result "$desc" "$rule" "$ima_policy" 2
desc="related actions, same func"
rule="measure func=FILE_CHECK"
ima_policy="dont_measure func=FILE_CHECK"
expect_pass check_result "$desc" "$rule" "$ima_policy" 2
desc="related actions, same func, different policy options"
rule="measure func=FILE_CHECK"
ima_policy="dont_measure func=FILE_CHECK permit_directio"
expect_pass check_result "$desc" "$rule" "$ima_policy" 2
desc="related actions, same func, different policy options"
rule="measure func=FILE_CHECK permit_directio"
ima_policy="dont_measure func=FILE_CHECK"
expect_pass check_result "$desc" "$rule" "$ima_policy" 2
desc="same actions, same func, same mask with different modifier (no disjoint sets with the ^ modifier)"
rule="measure func=FILE_CHECK mask=MAY_EXEC"
ima_policy="measure func=FILE_CHECK mask=^MAY_EXEC"
expect_pass check_result "$desc" "$rule" "$ima_policy" 2
desc="same actions, same func, different mask with same modifier (no disjoint sets with the ^ modifier)"
rule="measure func=FILE_CHECK mask=^MAY_READ"
ima_policy="measure func=FILE_CHECK mask=^MAY_EXEC"
expect_pass check_result "$desc" "$rule" "$ima_policy" 2
desc="same actions, same func, different policy options"
rule="measure func=FILE_CHECK"
ima_policy="measure func=FILE_CHECK permit_directio"
expect_pass check_result "$desc" "$rule" "$ima_policy" 2
desc="same actions, same func, different policy options"
rule="measure func=FILE_CHECK permit_directio"
ima_policy="measure func=FILE_CHECK"
expect_pass check_result "$desc" "$rule" "$ima_policy" 2
desc="same actions, MMAP_CHECK and MMAP_CHECK_REQPROT hooks"
rule="measure func=MMAP_CHECK"
ima_policy="measure func=MMAP_CHECK_REQPROT"
expect_pass check_result "$desc" "$rule" "$ima_policy" 2
desc="related actions, same func, same mask with same modifier"
rule="measure func=FILE_CHECK mask=^MAY_EXEC"
ima_policy="dont_measure func=FILE_CHECK mask=^MAY_EXEC"
expect_pass check_result "$desc" "$rule" "$ima_policy" 2
desc="same actions, same func, different uid with same operator (overlap because operators are not supported)"
rule="measure func=FILE_CHECK uid>0"
ima_policy="measure func=FILE_CHECK uid>1"
expect_pass check_result "$desc" "$rule" "$ima_policy" 2
desc="same actions, same func, same uid with different operator (overlap because operators are not supported)"
rule="measure func=FILE_CHECK uid>1"
ima_policy="measure func=FILE_CHECK uid<1"
expect_pass check_result "$desc" "$rule" "$ima_policy" 2
# Overlapping and same rules.
desc="same actions, same func"
rule="appraise func=FILE_CHECK"
ima_policy="appraise func=FILE_CHECK"
expect_pass check_result "$desc" "$rule" "$ima_policy" 4
desc="same actions, same func, same mask"
rule="appraise mask=MAY_READ func=FILE_CHECK"
ima_policy="appraise func=FILE_CHECK mask=MAY_READ"
expect_pass check_result "$desc" "$rule" "$ima_policy" 4
desc="same actions, same func, same mask, same policy options"
rule="appraise mask=MAY_READ func=FILE_CHECK permit_directio appraise_type=imasig"
ima_policy="appraise func=FILE_CHECK mask=MAY_READ permit_directio appraise_type=imasig"
expect_pass check_result "$desc" "$rule" "$ima_policy" 4
desc="same actions, same func"
rule="measure func=MMAP_CHECK_REQPROT"
ima_policy="measure func=MMAP_CHECK_REQPROT"
expect_pass check_result "$desc" "$rule" "$ima_policy" 4
desc="same actions, same func with alias (PATH_CHECK = FILE_CHECK)"
rule="measure func=FILE_CHECK"
ima_policy="measure func=PATH_CHECK"
expect_pass check_result "$desc" "$rule" "$ima_policy" 4
desc="same actions, same func with alias (PATH_CHECK = FILE_CHECK), same mask with same modifiers"
rule="measure mask=^MAY_READ func=FILE_CHECK"
ima_policy="measure func=PATH_CHECK mask=^MAY_READ"
expect_pass check_result "$desc" "$rule" "$ima_policy" 4
desc="same actions, same func with alias (PATH_CHECK = FILE_CHECK) and same mask with same modifiers, same uid with same operators"
rule="measure mask=^MAY_READ uid>0 func=FILE_CHECK"
ima_policy="measure func=PATH_CHECK mask=^MAY_READ uid>0"
expect_pass check_result "$desc" "$rule" "$ima_policy" 4
desc="same actions, same func with alias (PATH_CHECK = FILE_CHECK) and same mask with same modifiers, same uid with same operators"
rule="measure mask=^MAY_READ uid<1 func=FILE_CHECK"
ima_policy="measure func=PATH_CHECK mask=^MAY_READ uid<1"
expect_pass check_result "$desc" "$rule" "$ima_policy" 4
# Overlapping and two rules (one same, one different).
desc="first: same actions, same func, second: unrelated actions with appraise and a do action"
rule="appraise func=FILE_CHECK"
ima_policy="appraise func=FILE_CHECK\nmeasure func=FILE_CHECK"
expect_pass check_result "$desc" "$rule" "$ima_policy" 6
desc="first: unrelated actions with appraise and a do action, same func, second: same actions"
rule="appraise func=FILE_CHECK"
ima_policy="measure func=FILE_CHECK\nappraise func=FILE_CHECK"
expect_pass check_result "$desc" "$rule" "$ima_policy" 6
desc="first: same actions, same func, same mask, second: different policy options"
rule="appraise mask=MAY_READ func=FILE_CHECK"
ima_policy="appraise func=FILE_CHECK mask=MAY_READ\nappraise func=FILE_CHECK mask=MAY_READ permit_directio"
expect_pass check_result "$desc" "$rule" "$ima_policy" 6
desc="first: same actions, same func with alias (PATH_CHECK = FILE_CHECK), same mask, second: different policy options"
rule="appraise mask=MAY_READ func=FILE_CHECK"
ima_policy="appraise func=PATH_CHECK mask=MAY_READ\nappraise func=FILE_CHECK mask=MAY_READ permit_directio"
expect_pass check_result "$desc" "$rule" "$ima_policy" 6
# Non-overlapping and three rules.
desc="same actions, same func and mask, different uid"
rule="appraise mask=MAY_READ func=FILE_CHECK uid=0"
ima_policy="appraise mask=MAY_READ func=FILE_CHECK uid=1\nappraise mask=MAY_READ func=FILE_CHECK uid=2\nappraise mask=MAY_READ func=FILE_CHECK uid=3"
expect_pass check_result "$desc" "$rule" "$ima_policy" 0
desc="same actions, same func and mask, different uid, except one that is the same"
rule="appraise mask=MAY_READ func=FILE_CHECK uid=0"
ima_policy="appraise mask=MAY_READ func=FILE_CHECK uid=1\nappraise mask=MAY_READ func=FILE_CHECK uid=0\nappraise mask=MAY_READ func=FILE_CHECK uid=3"
expect_pass check_result "$desc" "$rule" "$ima_policy" 4

6
tests/install-fsverity.sh Executable file
View File

@ -0,0 +1,6 @@
#!/bin/sh
git clone https://git.kernel.org/pub/scm/fs/fsverity/fsverity-utils.git
cd fsverity-utils
CC=gcc make -j$(nproc)
cd ..

View File

@ -0,0 +1,6 @@
#!/bin/sh
git clone https://github.com/brauner/mount-idmapped.git
cd mount-idmapped
gcc -o mount-idmapped mount-idmapped.c
cd ..

View File

@ -13,7 +13,14 @@ wget --no-check-certificate https://github.com/openssl/openssl/archive/refs/tags
tar --no-same-owner -xzf ${version}.tar.gz
cd openssl-${version}
./Configure --prefix=/opt/openssl3 --openssldir=/opt/openssl3/ssl
if [ "$VARIANT" = "i386" ]; then
echo "32-bit compilation"
FLAGS="-m32 linux-generic32"
fi
./Configure $FLAGS no-engine no-dynamic-engine --prefix=/opt/openssl3 --openssldir=/opt/openssl3
# Uncomment for debugging
# perl configdata.pm --dump | grep engine
make -j$(nproc)
# only install apps and library
sudo make install_sw

View File

@ -9,7 +9,7 @@ else
SUDO=sudo
fi
version=1637
version=1682
wget --no-check-certificate https://sourceforge.net/projects/ibmswtpm2/files/ibmtpm${version}.tar.gz/download
mkdir ibmtpm$version

407
tests/mmap_check.test Executable file
View File

@ -0,0 +1,407 @@
#!/bin/bash
# SPDX-License-Identifier: GPL-2.0
#
# Copyright (C) 2022-2023 Roberto Sassu <roberto.sassu@huawei.com>
#
# Check the behavior of MMAP_CHECK and MMAP_CHECK_REQPROT
trap '_report_exit_and_cleanup _cleanup_env cleanup' SIGINT SIGTERM SIGSEGV EXIT
PATCHES=(
'ima: Align ima_file_mmap() parameters with mmap_file LSM hook'
'ima: Introduce MMAP_CHECK_REQPROT hook'
)
RET_INVALID_RULE=$((0x0001))
RET_RULE_OVERLAP=$((0x0002))
RET_SAME_RULE_EXISTS=$((0x0004))
EVM_INIT_HMAC=$((0x0001))
EVM_INIT_X509=$((0x0002))
# Base VERBOSE on the environment variable, if set.
VERBOSE="${VERBOSE:-0}"
# Errors defined in test_mmap
ERR_SETUP=1
ERR_TEST=2
cd "$(dirname "$0")" || exit 1
export PATH=$PWD/../src:$PWD:$PATH
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH
. ./functions.sh
_require evmctl
cleanup() {
if [ "$g_loop_mounted" = "1" ]; then
popd > /dev/null || exit "$FAIL"
umount "$g_mountpoint"
fi
if [ -n "$g_dev" ]; then
losetup -d "$g_dev"
fi
if [ -n "$g_image" ]; then
rm -f "$g_image"
fi
if [ -n "$g_mountpoint" ]; then
rm -Rf "$g_mountpoint"
fi
if [ -n "$g_key_path_der" ]; then
rm -f "$g_key_path_der"
fi
}
# Use the fsuuid= IMA policy keyword to select only files created/used by the
# tests below. Also use fowner= to differentiate between files created/used by
# individual tests.
IMA_UUID="28b23254-9467-44c0-b6ba-34b12e85a26e"
MEASURE_MMAP_CHECK_FOWNER=2000
MEASURE_MMAP_CHECK_REQPROT_FOWNER=2001
MEASURE_MMAP_CHECK_RULE="measure func=MMAP_CHECK fsmagic=0xef53 fsuuid=$IMA_UUID fowner=$MEASURE_MMAP_CHECK_FOWNER"
MEASURE_MMAP_CHECK_REQPROT_RULE="measure func=MMAP_CHECK_REQPROT fsmagic=0xef53 fsuuid=$IMA_UUID fowner=$MEASURE_MMAP_CHECK_REQPROT_FOWNER"
APPRAISE_MMAP_CHECK_FOWNER=2002
APPRAISE_MMAP_CHECK_REQPROT_FOWNER=2003
APPRAISE_MMAP_CHECK_RULE="appraise func=MMAP_CHECK fsmagic=0xef53 fsuuid=$IMA_UUID fowner=$APPRAISE_MMAP_CHECK_FOWNER"
APPRAISE_MMAP_CHECK_REQPROT_RULE="appraise func=MMAP_CHECK_REQPROT fsmagic=0xef53 fsuuid=$IMA_UUID fowner=$APPRAISE_MMAP_CHECK_REQPROT_FOWNER"
check_load_ima_rule() {
local result new_policy color
echo -e "$1\n$(cat /sys/kernel/security/ima/policy)" | ima_policy_check.awk
result=$?
if [ $((result & RET_INVALID_RULE)) -eq $RET_INVALID_RULE ]; then
echo "${RED}Invalid rule${NORM}"
return "$HARDFAIL"
fi
if [ $((result & RET_RULE_OVERLAP)) -eq $RET_RULE_OVERLAP ]; then
color=${YELLOW}
if [ -n "$TST_ENV" ]; then
color=${RED}
fi
echo "${color}Possible interference with existing IMA policy rule${NORM}"
if [ -n "$TST_ENV" ]; then
return "$HARDFAIL"
fi
fi
if [ $((result & RET_SAME_RULE_EXISTS)) -eq $RET_SAME_RULE_EXISTS ]; then
return "$OK"
fi
new_policy=$(mktemp -p "$g_mountpoint")
echo "$1" > "$new_policy"
echo "$new_policy" > /sys/kernel/security/ima/policy
result=$?
rm -f "$new_policy"
if [ "$result" -ne 0 ]; then
echo "${RED}Failed to set IMA policy${NORM}"
return "$HARDFAIL"
fi
return "$OK"
}
check_mmap() {
local hook="$1"
local arg="$2"
local test_file fowner rule result test_file_entry
echo -e "\nTest: ${FUNCNAME[0]} (hook=\"$hook\", test_mmap arg: \"$arg\")"
if ! test_file=$(mktemp -p "$PWD"); then
echo "${RED}Cannot create $test_file${NORM}"
return "$HARDFAIL"
fi
if ! echo "test" > "$test_file"; then
echo "${RED}Cannot write $test_file${NORM}"
return "$FAIL"
fi
fowner="$MEASURE_MMAP_CHECK_FOWNER"
rule="$MEASURE_MMAP_CHECK_RULE"
if [ "$hook" = "MMAP_CHECK_REQPROT" ]; then
fowner="$MEASURE_MMAP_CHECK_REQPROT_FOWNER"
rule="$MEASURE_MMAP_CHECK_REQPROT_RULE"
fi
if ! chown "$fowner" "$test_file"; then
echo "${RED}Cannot change owner of $test_file${NORM}"
return "$HARDFAIL"
fi
check_load_ima_rule "$rule"
result=$?
if [ $result -ne "$OK" ]; then
return $result
fi
test_mmap "$test_file" "$arg"
result=$?
if [ $result -ne 0 ] && [ $result -ne "$ERR_TEST" ]; then
echo "${RED}Unexpected exit status $result from test_mmap${NORM}"
return "$HARDFAIL"
fi
if [ "$TFAIL" != "yes" ]; then
echo -n "Result (expect found): "
else
echo -n "Result (expect not found): "
fi
test_file_entry=$(awk '$5 == "'"$test_file"'"' < /sys/kernel/security/ima/ascii_runtime_measurements)
if [ -z "$test_file_entry" ]; then
if [ "$TFAIL" != "yes" ]; then
echo "${RED}not found${NORM}"
else
echo "${GREEN}not found${NORM}"
fi
return "$FAIL"
fi
if [ "$TFAIL" != "yes" ]; then
echo "${GREEN}found${NORM}"
else
echo "${RED}found${NORM}"
fi
if [ "$VERBOSE" -gt 0 ]; then
echo "$test_file_entry"
fi
return "$OK"
}
check_deny() {
local hook="$1"
local arg="$2"
local test_file fowner rule result
echo -e "\nTest: ${FUNCNAME[0]} (hook=\"$hook\", test_mmap arg: \"$arg\")"
if ! test_file=$(mktemp -p "$PWD"); then
echo "${RED}Cannot create $test_file${NORM}"
return "$HARDFAIL"
fi
if ! echo "test" > "$test_file"; then
echo "${RED}Cannot write $test_file${NORM}"
return "$FAIL"
fi
if ! evmctl ima_sign -a sha256 --key "$g_key_path" "$test_file" &> /dev/null; then
echo "${RED}Cannot sign $test_file${NORM}"
return "$HARDFAIL"
fi
fowner="$APPRAISE_MMAP_CHECK_FOWNER"
rule="$APPRAISE_MMAP_CHECK_RULE"
if [ "$hook" = "MMAP_CHECK_REQPROT" ]; then
fowner="$APPRAISE_MMAP_CHECK_REQPROT_FOWNER"
rule="$APPRAISE_MMAP_CHECK_REQPROT_RULE"
fi
if ! chown "$fowner" "$test_file"; then
echo "${RED}Cannot change owner of $test_file${NORM}"
return "$HARDFAIL"
fi
check_load_ima_rule "$rule"
result=$?
if [ $result -ne "$OK" ]; then
return $result
fi
test_mmap "$test_file" exec
result=$?
if [ $result -ne 0 ] && [ $result -ne "$ERR_TEST" ]; then
echo "${RED}Unexpected exit status $result from test_mmap${NORM}"
return "$HARDFAIL"
fi
test_mmap "$test_file" "$arg"
result=$?
if [ $result -ne 0 ] && [ $result -ne "$ERR_TEST" ]; then
echo "${RED}Unexpected exit status $result from test_mmap${NORM}"
return "$HARDFAIL"
fi
if [ "$TFAIL" != "yes" ]; then
echo -n "Result (expect denied): "
else
echo -n "Result (expect allowed): "
fi
if [ $result -eq 0 ]; then
if [ "$TFAIL" != "yes" ]; then
echo "${RED}allowed${NORM}"
else
echo "${GREEN}allowed${NORM}"
fi
return "$FAIL"
fi
if [ "$TFAIL" != "yes" ]; then
echo "${GREEN}denied${NORM}"
else
echo "${RED}denied${NORM}"
fi
return "$OK"
}
# Run in the new environment if TST_ENV is set.
_run_env "$TST_KERNEL" "$PWD/$(basename "$0")" "TST_ENV=$TST_ENV TST_KERNEL=$TST_KERNEL PATH=$PATH LD_LIBRARY_PATH=$LD_LIBRARY_PATH VERBOSE=$VERBOSE TST_KEY_PATH=$TST_KEY_PATH"
# Exit from the creator of the new environment.
_exit_env "$TST_KERNEL"
# Mount filesystems in the new environment.
_init_env
if [ "$(whoami)" != "root" ]; then
echo "${CYAN}This script must be executed as root${NORM}"
exit "$SKIP"
fi
if [ ! -f /sys/kernel/security/ima/policy ]; then
echo "${CYAN}IMA policy file not found${NORM}"
exit "$SKIP"
fi
if ! cat /sys/kernel/security/ima/policy &> /dev/null; then
echo "${CYAN}IMA policy file is not readable${NORM}"
exit "$SKIP"
fi
if [ -n "$TST_KEY_PATH" ]; then
if [ "${TST_KEY_PATH:0:1}" != "/" ]; then
echo "${RED}Absolute path required for the signing key${NORM}"
exit "$FAIL"
fi
if [ ! -f "$TST_KEY_PATH" ]; then
echo "${RED}Kernel signing key not found in $TST_KEY_PATH${NORM}"
exit "$FAIL"
fi
g_key_path="$TST_KEY_PATH"
elif [ -f "$PWD/../signing_key.pem" ]; then
g_key_path="$PWD/../signing_key.pem"
elif [ -f "/lib/modules/$(uname -r)/source/certs/signing_key.pem" ]; then
g_key_path="/lib/modules/$(uname -r)/source/certs/signing_key.pem"
elif [ -f "/lib/modules/$(uname -r)/build/certs/signing_key.pem" ]; then
g_key_path="/lib/modules/$(uname -r)/build/certs/signing_key.pem"
else
echo "${CYAN}Kernel signing key not found${NORM}"
exit "$SKIP"
fi
evm_value=$(cat /sys/kernel/security/evm)
if [ $((evm_value & EVM_INIT_X509)) -eq "$EVM_INIT_X509" ]; then
if [ $((evm_value & EVM_INIT_HMAC)) -ne "$EVM_INIT_HMAC" ]; then
echo "${CYAN}Incompatible EVM mode $evm_value${NORM}"
exit "$SKIP"
fi
fi
g_key_path_der=$(mktemp)
openssl x509 -in "$g_key_path" -out "$g_key_path_der" -outform der
if ! keyctl padd asymmetric pubkey %keyring:.ima < "$g_key_path_der" &> /dev/null; then
echo "${RED}Public key cannot be added to the IMA keyring${NORM}"
exit "$FAIL"
fi
g_mountpoint=$(mktemp -d)
g_image=$(mktemp)
if [ -z "$g_mountpoint" ]; then
echo "${RED}Mountpoint directory not created${NORM}"
exit "$FAIL"
fi
if ! dd if=/dev/zero of="$g_image" bs=1M count=20 &> /dev/null; then
echo "${RED}Cannot create test image${NORM}"
exit "$FAIL"
fi
g_dev=$(losetup -f "$g_image" --show)
if [ -z "$g_dev" ]; then
echo "${RED}Cannot create loop device${NORM}"
exit "$FAIL"
fi
if ! mkfs.ext4 -U "$IMA_UUID" -b 4096 "$g_dev" &> /dev/null; then
echo "${RED}Cannot format $g_dev${NORM}"
exit "$FAIL"
fi
if ! mount -o iversion "$g_dev" "$g_mountpoint"; then
echo "${RED}Cannot mount loop device${NORM}"
exit "$FAIL"
fi
g_loop_mounted=1
pushd "$g_mountpoint" > /dev/null || exit "$FAIL"
# Ensure that IMA does not add a new measurement entry if an application calls
# mmap() with PROT_READ, and a policy rule contains the MMAP_CHECK hook.
# In this case, both the protections requested by the application and the final
# protections applied by the kernel contain only PROT_READ, so there is no
# match with the IMA rule, which expects PROT_EXEC to be set.
expect_fail check_mmap "MMAP_CHECK" ""
# Ensure that IMA adds a new measurement entry if an application calls mmap()
# with PROT_READ | PROT_EXEC, and a policy rule contains the MMAP_CHECK hook.
expect_pass check_mmap "MMAP_CHECK" "exec"
# Same as in the first test, but in this case the application calls the
# personality() system call with READ_IMPLIES_EXEC, which causes the kernel to
# add PROT_EXEC in the final protections passed to the MMAP_CHECK hook.
#
# Ensure that the bug introduced by 98de59bfe4b2 ("take calculation of final
# protections in security_mmap_file() into a helper") is fixed, by passing the
# final protections again to the MMAP_CHECK hook. Due to the bug, the hook
# received the protections requested by the application. Since those protections
# don't have PROT_EXEC, IMA was not creating a measurement entry.
expect_pass_if '0' check_mmap "MMAP_CHECK" "read_implies_exec"
# Repeat the previous three tests, but with the new MMAP_CHECK_REQPROT hook,
# which behaves like the buggy MMAP_CHECK hook. In the third test, expect that
# no new measurement entry is created, since the MMAP_CHECK_REQPROT hook sees
# the protections requested by the application (PROT_READ).
expect_fail_if '1' check_mmap "MMAP_CHECK_REQPROT" ""
expect_pass_if '1' check_mmap "MMAP_CHECK_REQPROT" "exec"
expect_fail_if '1' check_mmap "MMAP_CHECK_REQPROT" "read_implies_exec"
# Ensure that IMA refuses an mprotect() with PROT_EXEC on a memory area
# obtained with an mmap() with PROT_READ. This is due to the inability of IMA
# to measure/appraise the file for which mmap() was called (locking issue).
expect_pass check_deny "MMAP_CHECK" "mprotect"
# Ensure that MMAP_CHECK_REQPROT has the same behavior of MMAP_CHECK for the
# previous test.
expect_pass_if '1' check_deny "MMAP_CHECK_REQPROT" "mprotect"
# Ensure that there cannot be an mmap() with PROT_EXEC on a file with writable
# mappings, due to the inability of IMA to make a reliable measurement of that
# file.
expect_pass check_deny "MMAP_CHECK" "exec_on_writable"
# Ensure that MMAP_CHECK_REQPROT has the same behavior of MMAP_CHECK for the
# previous test.
expect_pass_if '1' check_deny "MMAP_CHECK_REQPROT" "exec_on_writable"

1122
tests/portable_signatures.test Executable file

File diff suppressed because it is too large Load Diff

View File

@ -17,6 +17,10 @@
cd "$(dirname "$0")" || exit 1
PATH=../src:$PATH
# set the env SIGV1=1 to execute the signature v1 tests
SIGV1=${SIGV1:-0}
source ./functions.sh
_require cmp evmctl getfattr openssl xxd
@ -368,13 +372,18 @@ try_different_sigs() {
## Test v1 signatures
# Signature v1 only supports sha1 and sha256 so any other should fail
expect_fail \
if [ $SIGV1 -eq 0 ]; then
__skip() { echo "IMA signature v1 tests are skipped: not supported"; return $SKIP; }
expect_pass __skip
else
expect_fail \
check_sign TYPE=ima KEY=rsa1024 ALG=md5 PREFIX=0x0301 OPTS=--rsa
sign_verify rsa1024 sha1 0x0301 --rsa
sign_verify rsa1024 sha256 0x0301 --rsa
sign_verify rsa1024 sha1 0x0301 --rsa
sign_verify rsa1024 sha256 0x0301 --rsa
try_different_keys
try_different_sigs
fi
## Test v2 signatures with RSA PKCS#1
# List of allowed hashes much greater but not all are supported.
@ -407,9 +416,12 @@ sign_verify prime256v1 sha384 0x030205:K:004[345678]
sign_verify prime256v1 sha512 0x030206:K:004[345678]
# If openssl 3.0 is installed, test the SM2/3 algorithm combination
if [ -x /opt/openssl3/bin/openssl ]; then
PATH=/opt/openssl3/bin:$PATH LD_LIBRARY_PATH=/opt/openssl3/lib \
ssl_major_version=$(openssl version | sed -n 's/^OpenSSL \([^\.]\).*/\1/p')
if [ "${ssl_major_version}" = 3 ]; then
sign_verify sm2 sm3 0x030211:K:004[345678]
else
__skip() { echo "sm2/sm3 tests are skipped (ssl version)"; return $SKIP; }
expect_pass __skip
fi
# Test v2 signatures with EC-RDSA

128
tests/test_mmap.c Normal file
View File

@ -0,0 +1,128 @@
// SPDX-License-Identifier: GPL-2.0
/*
* Copyright (C) 2023 Huawei Technologies Duesseldorf GmbH
*
* Tool to test IMA MMAP_CHECK and MMAP_CHECK_REQPROT hooks.
*/
#include <stdio.h>
#include <errno.h>
#include <fcntl.h>
#include <string.h>
#include <unistd.h>
#include <sys/stat.h>
#include <sys/mman.h>
#include <sys/personality.h>
/*
* Convention: return 1 for errors that should not occur, as they are
* setup-related, return 2 for errors that might occur due to testing
* conditions.
*/
#define ERR_SETUP 1
#define ERR_TEST 2
int main(int argc, char *argv[])
{
struct stat st;
void *ptr, *ptr_write = NULL;
int ret, fd, fd_write, prot = PROT_READ;
if (!argv[1]) {
printf("Missing file parameter\n");
return ERR_SETUP;
}
if (argv[2] && !strcmp(argv[2], "read_implies_exec")) {
ret = personality(READ_IMPLIES_EXEC);
if (ret == -1) {
printf("Failed to set personality, err: %d (%s)\n",
-errno, strerror(errno));
return ERR_SETUP;
}
}
if (stat(argv[1], &st) == -1) {
printf("Failed to access %s, err: %d (%s)\n", argv[1], -errno,
strerror(errno));
return ERR_SETUP;
}
if (argv[2] && !strcmp(argv[2], "exec_on_writable")) {
fd_write = open(argv[1], O_RDWR);
if (fd_write == -1) {
printf("Failed to open %s in r/w, err: %d (%s)\n",
argv[1], -errno, strerror(errno));
return ERR_SETUP;
}
ptr_write = mmap(0, st.st_size, PROT_WRITE, MAP_SHARED,
fd_write, 0);
close(fd_write);
if (ptr_write == MAP_FAILED) {
printf("Failed mmap() with PROT_WRITE on %s, err: %d (%s)\n",
argv[1], -errno, strerror(errno));
return ERR_SETUP;
}
}
fd = open(argv[1], O_RDONLY);
if (fd == -1) {
printf("Failed to open %s in ro, err: %d (%s)\n", argv[1],
-errno, strerror(errno));
if (ptr_write && munmap(ptr_write, st.st_size) == -1)
printf("Failed munmap() of writable mapping on %s, err: %d (%s)\n",
argv[1], -errno, strerror(errno));
return ERR_SETUP;
}
if (argv[2] && !strncmp(argv[2], "exec", 4))
prot |= PROT_EXEC;
ptr = mmap(0, st.st_size, prot, MAP_PRIVATE, fd, 0);
close(fd);
if (ptr_write && munmap(ptr_write, st.st_size) == -1) {
printf("Failed munmap() of writable mapping on %s, err: %d (%s)\n",
argv[1], -errno, strerror(errno));
return ERR_SETUP;
}
if (ptr == MAP_FAILED) {
ret = ERR_SETUP;
if (argv[2] && !strcmp(argv[2], "exec_on_writable") &&
errno == EACCES)
ret = ERR_TEST;
else
printf("Failed mmap() with PROT_READ%s on %s, err: %d (%s)\n",
(prot & PROT_EXEC) ? " | PROT_EXEC" : "",
argv[1], -errno, strerror(errno));
return ret;
}
ret = 0;
if (argv[2] && !strcmp(argv[2], "mprotect")) {
ret = mprotect(ptr, st.st_size, PROT_EXEC);
if (ret == -1) {
ret = ERR_SETUP;
if (errno == EPERM)
ret = ERR_TEST;
else
printf("Unexpected mprotect() error on %s, err: %d (%s)\n",
argv[1], -errno, strerror(errno));
}
}
if (munmap(ptr, st.st_size) == -1) {
printf("Failed munmap() of mapping on %s, err: %d (%s)\n",
argv[1], -errno, strerror(errno));
return ERR_SETUP;
}
return ret;
}