The TPM2_NV_Read commands returns the requested data in a
TPM2B_MAX_NV_BUFFER type, the maximum size of which is TPM-specific and
can be determined by querying the value of the TPM_PT_NV_BUFFER_MAX
property.
The value of this may be smaller than the payload size of some NV
indexes, particularly when that payload is a X509 certificate with a RSA
public key. Eg, the manufacturer supplied RSA EK certificate on my own
machine has a size of 1035 bytes, and the value of TPM_PT_NV_BUFFER_MAX
is 1024.
To handle this case and make it possible to read any EK certificate from
the TPM, make tpm2_read_nv_index support chunked reads when the payload
size is larger than what the TPM can return in a single command.
Cover the full preference order (SHA256 > SHA384 > SHA512 > SHA1), the
single-bank cases and the -EOPNOTSUPP path. Also test the _legacy()
variant: a set with SHA384/SHA512 active but no SHA256 must resolve to
SHA1, not SHA384, pinning down the backwards-compatibility behavior that
keeps legacy sealed secrets unsealable.
Co-developed-by: Claude Opus 4.8 <noreply@anthropic.com>
Until now OpenSSL was linked into every binary and library that needed
cryptography, pulling libcrypto (and, for resolved, libssl) into the
address space of services that never touch them at runtime. This commit
moves all OpenSSL usage behind the same dlopen helper pattern that we
already use for other optional libraries (libpam, libseccomp, libxz, …)
so libcrypto/libssl are only loaded on demand.
The bulk of the work lives in src/shared/crypto-util.{c,h} (libcrypto)
and src/shared/ssl-util.{c,h} (libssl), which replace the previous
src/shared/openssl-util.{c,h}:
- crypto-util.{c,h} declares every libcrypto function we call via
DLSYM_PROTOTYPE() and resolves them inside dlopen_libcrypto().
- ssl-util.{c,h} holds the libssl-specific DLSYM_PROTOTYPEs,
dlopen_libssl(), and the SSL_freep cleanup helper, so translation
units that only need libcrypto do not pull in libssl declarations.
- Callers refer to the symbols through sym_* aliases rather than the
original names.
- Convenience macros that used to be provided by the OpenSSL headers
(OPENSSL_free, BN_num_bytes, the sk_TYPE_* helpers, …) are
reimplemented as sym_* wrappers so no code path needs to fall back
to the linker-resolved symbols.
- All _cleanup_ helpers are redefined in terms of the sym_* variants
(EVP_PKEY_freep, X509_freep, BIO_freep, …) so cleanup attributes
keep working without pulling in libcrypto symbols at link time.
- The public crypto-util.c entry points (openssl_pubkey_from_pem,
openssl_digest_many, openssl_hmac_many, openssl_cipher_many,
kdf_ss_derive, kdf_kb_hmac_derive, rsa_* / ecc_* helpers,
pubkey_fingerprint, digest_and_sign, pkcs7_new, x509_fingerprint,
openssl_extract_public_key, pkey_generate_volume_keys, the load_*
helpers, …) now call dlopen_libcrypto() at entry before touching any
sym_* pointer.
The call sites across the tree have been converted to call
dlopen_libcrypto()/dlopen_libssl() at the appropriate entry point
before their first sym_* use, and to use sym_* variants throughout:
- bootctl, sbsign, measure, pcrlock, pcrextend, tpm2-setup, repart,
cryptsetup, cryptenroll, homectl, homed, homework, keyutil,
sysupdate, creds, import, dissect-image, pe-binary, pkcs11-util,
pkcs7-util, tpm2-util, creds-util. resolved additionally dlopens
libssl for DoT.
The meson build files are updated to depend on libopenssl_cflags (a
new partial dependency that exposes include paths and compile flags
only, not the linker flags) instead of libopenssl for every target that
previously linked against OpenSSL. Nothing links against libcrypto or
libssl directly anymore.
A new src/sbsign/authenticode.c hosts the Authenticode ASN.1 type
definitions that used to live inline in sbsign.c. The OpenSSL
ASN1_SEQUENCE / ASN1_CHOICE / IMPLEMENT_ASN1_FUNCTIONS macros expand to
code that references libcrypto symbols directly, so to keep this
translation unit unlinked from libcrypto we redirect ASN1_item_* to
the sym_* variants via #define and wrap the ASN1_*_it() getters (which
appear as constant function pointers in static initializers) in small
trampoline functions that forward to the sym_* pointers at runtime.
test-dlopen-so gains assertions for dlopen_libcrypto and dlopen_libssl
so the dlopen contract is exercised in CI, and the openssl-specific
test was renamed from test-openssl.c to test-crypto-util.c to match
the new header naming.
All major distributions have switched to OpenSSL version 3.x.
Let's drop support of OpenSSL version 1.x.
Note, OpenSSL 3.0 was released on 2021-09-07 (and will be EOL on 2026-09-07).
See also #38608.
To make sure everything still compiles, we add a preliminary include
of forward.h to tests.h to make sure it is included in every test source
file. We'll clean up the tests.h includes in a later commit.
We also add a <errno.h> include to errno-list.h to keep test-errno-list.c
compiling. It'll be removed again when we clean up includes in src/basic.
Split out of #37344.
tpm2_parse_pcr_argument_to_mask() is supposed to parse a PCR mask
string, and uses the full blown tpm2_parse_pcr_argument() call at its
core, which parses more than just a mask, i.e. values and algorithms
too. Which is very confusing at times, because commands such as
"systemd-cryptenroll --tpm2-device=auto
--tpm2-public-key-pcrs=1:sha1=09dbdbc7f6cdd8029cc90b57a915c19a0ac21bce"
are very confusing, since they suggest enrollment with a specific
algorithm and has value, but this is not in fact what happens: both are
entirely ignored.
That this was accepted this way was more an accident than intended,
which is already visible in the fact that extensive test case entirely
ignores the fact that strings like this are accepted.
So far you had to pick:
1. Use a signed PCR TPM2 policy to lock your disk to (i.e. UKI vendor
blesses your setup via signature)
or
2. Use a pcrlock policy (i.e. local system blesses your setup via
dynamic local policy stored in NV index)
It was not possible combine these two, because TPM2 access policies do
not allow the combination of PolicyAuthorize (used to implement #1
above) and PolicyAuthorizeNV (used to implement #2) in a single policy,
unless one is "further upstream" (and can simply remove the other from
the policy freely).
This is quite limiting of course, since we actually do want to enforce
on each TPM object that both the OS vendor policy and the local policy
must be fulfilled, without the chance for the vendor or the local system
to disable the other.
This patch addresses this: instead of trying to find a way to come up
with some adventurous scheme to combine both policy into one TPM2
policy, we simply shard the symmetric LUKS decryption key: one half we
protect via the signed PCR policy, and the other we protect via the
pcrlock policy. Only if both halves can be acquired the disk can be
decrypted.
This means:
1. we simply double the unlock key in length in case both policies shall
be used.
2. We store two resulting TPM policy hashes in the LUKS token JSON, one
for each policy
3. We store two sealed TPM policy key blobs in the LUKS token JSON, for
both halves of the LUKS unlock key.
This patch keeps the "sharding" logic relatively generic (i.e. the low
level logic is actually fine with more than 2 shards), because I figure
sooner or later we might have to encode more shards, for example if we
add further TPM2-based access policies, for example when combining FIDO2
with TPM2, or implementing TOTP for this.
Use FOREACH_ELEMENT where possible. Generated with this command,
and checked manually:
git grep -l 'FOREACH_ARRAY.*ELEMENTSOF' | \
xargs sed -ri 's/FOREACH_ARRAY\((.*), (.*), (ELEMENTSOF.*)\)/FOREACH_ELEMENT(\1, \2)/'
Check the calculated TPM2B_PUBLIC key "name" to verify our PEM->TPM2B_PUBLIC
function remains consistent with previous code. This is important as the
TPM2B_PUBLIC "name" is used in the Authorize policy and so any change to a key
"name" would break unsealing for previously-sealed objects (see bug #30546).
Note that the tpm2_tpm2b_public_from_openssl_pkey() function results in a
TPM2B_PUBLIC with the same "name" as using the tpm2-tools program
tpm2_loadexternal, at least as of tpm2-tools version 5.6.18, with the test keys
from TEST(tpm2b_public_from_openssl_pkey) in src/test/test-tpm2.
The tpm2 test currently verifies PEM->TPM2B_PUBLIC conversion for an RSA key
with the default exponent (0x10001); this adds verification for an RSA key with
a non-default exponent.
The openssl default value for an RSA key exponent value is 0x10001, and the TPM
specification defines a exponent value of 0 as representing this value. The
systemd code that converted an RSA PEM public key to a TPM2B_PUBLIC object
previously used the exponent value directly, but commit
e3acb4d24c changed the conversion to use the
special case exponent value of 0 for any RSA key with an exponent value of
0x10001.
Because the entire TPM2B_PUBLIC object is used to calculate its "name", this
difference in exponent value (0x10001 vs 0) introduced a change in the key
"name". Since the Authorize policy uses the key "name" directly in its policy
session hash value, this change resulted in new systemd code being unable to
properly unseal any data (e.g. a LUKS volume) that was previously sealed.
This reverts the code to no longer override an RSA exponent value of 0x10001
with the special case value of 0.
Fixes a bug introduced by commit e3acb4d24c.
Fixes: #30546
The TPM2 tests that genreate an RSA primary key are fast on vtpms, but
very slow on physical TPMs, simply because TPMs aren't precisely fast
devices. It makes sense to keep the tests around however. Hence hide the
test behind the "slow test" logic by default – but only if we run on
physical hw, and keep them in place on VMs (where we'd expect a vtpm, if
any).
The test-tpm2 test multiplexes a bunch of tests from a single
entrypoint test that creates the TPM2 connection. This means we only get
the nice log output which test we are looking for once for the
entrypoint.
Let's add a macro that allows it to nicely generate it for the inner
tests too and use it.
The SRK templates are defined by specification, so move the check for TPM
support to the tpm2_get_best_srk_template() function, and allow anyone to get
the ECC and RSA templates.
Also add test to verify the SRK templates are correct.
We'll soon have a function for determining the name of an NV index,
hence let's rename the existing function for the same of a public key to
make clear it's about public keys only.
In TEST-70-TPM2, test systemd-cryptenroll --tpm2-seal-key-handle using the
default (0) as well as the SRK handle (0x81000001), and test using a non-SRK
handle index after creating and persisting a primary key.
In test/test-tpm2, test tpm2_seal() and tpm2_unseal() using default (0), the SRK
handle, and a transient handle.
The test expects TPM2_CC_FIRST - 1 and TPM2_CC_LAST + 1 to be unsupported, but
those are not necessarily invalid commands. Instead test known-invalid
commands. Also add some more valid commands.
We so far maintained two places for symboic names for PCRs. One in
tpm2-util.h and one in tpm-pcr.h.
Let's unify this into one, i.e. move the full list from tpm2-util.h into
tpm-pcr.h, replacing the short list placed so far there.
Systematically prefix the definitions with TPM2_ or tpm2_, to follow how
we do this for all other defines in this context.
No change in behaviour, just unification of tables.
Add functions to convert a PEM or pkey to TPM2B_PUBLIC, and functions to
convert TPM2B_PUBLIC to pkey or fingerprint.
Supports both RSA and ECC keys.
Add ECC support to some test-tpm2 tests, and tests to cover the newly added functions.
In order to allow users to specify expected PCR values, change the
tpm2_parse_pcr_argument() to parse the text argument into an array of
Tpm2PCRValue objects, which provide not only the selected PCR indexes, but also
(optionally) the hash algorithm and hash value for each PCR index.
An array of Tpm2PCRValue objects effectively replaces a TPML_PCR_SELECTION
object combined with an array of (properly ordered) TPM2B_DIGEST objects.
Also update tpm2_calculate_sealing_policy() pin parameter to boolean use_pin,
since the function does not need to know the pin value, only if a pin is being
used.
This test loops through masks, but is a relatively long test due to the
increment size between loops; this slightly increases the increment size (from
3->5) which greatly speeds up the test.
Add functions to get either the 'legacy' or standard SRK template, for RSA or
ECC. The 'legacy' templates are those used with earlier code, where a transient
key was created to use for tpm sealing; the standard SRK is the persistent
shared key as defined in TCG guidance docs.
This also replaces tpm2_get_primary_template() with the new functions; that
function's use of flags is confusing and unnecessary.
Remove _cleanup_tpm2_context_ and _cleanup_tpm2_handle_ macros, replacing their
use with _cleanup_(tpm2_context_unrefp) and _cleanup_(tpm2_handle_freep),
respectively.
This adds tpm2_get_capability_algs(), tpm2_supports_alg(), and
tpm2_test_parms(). These functions allow verifying that the TPM supports
specific algs and parameters.
When creating a new context, this checks if the TPM supports the symmetric algs
we use. If the TPM does not support the symmetric algs and parameters we
require, we log and return error.
This adds functions to get the digest for a PolicyAuthorize operation. For
building a policy hash, this provides a function to calculate the hash; and for
building a policy hash to satisfy the authPolicy for an existing object, this
provides a function to perform PolicyAuthorize with an existing session.
This adds functions to get the digest for a PolicyAuthValue operation. For
building a policy hash, this provides a function to calculate the hash; and for
building a policy hash to satisfy the authPolicy for an existing object, this
provides a function to perform PolicyAuthValue with an existing session.
This adds functions to get the digest for a PolicyPCR operation. For building
a policy hash, this provides a function to calculate the hash; and for building
a policy hash to satisfy the authPolicy for an existing object, this provides a
function to perform PolicyPCR with an existing session.
Chris Coulson
Ivan Kruglov
Yu Watanabe
Daan De Meyer
Daan De Meyer
Luca Boccassi
Jelle van der Waa
Lennart Poettering
Integral
Matteo Croce
Dan Streetman