`gcc-14` added a new `-Wcalloc-transposed-args` warning recently. It
detected minor infelicity in `calloc()` API usage in `mbedtls`:
In file included from /build/mbedtls/tests/include/test/ssl_helpers.h:19,
from /build/mbedtls/tests/src/test_helpers/ssl_helpers.c:11:
/build/mbedtls/tests/src/test_helpers/ssl_helpers.c: In function 'mbedtls_test_init_handshake_options':
/build/mbedtls/tests/include/test/macros.h:128:46:
error: 'calloc' sizes specified with 'sizeof' in the earlier argument
and not in the later argument [-Werror=calloc-transposed-args]
128 | (pointer) = mbedtls_calloc(sizeof(*(pointer)), \
| ^
Signed-off-by: Sergei Trofimovich <slyich@gmail.com>
In principle the RSA tests shouldn't depend on RSA key generation: they just
need to operate on RSA keys. However they do need some method of creating an
RSA key, and we're currently doing random generation. So depend on what the
test code needs.
Depend on the legacy RSA interface, since driver-only RSA isn't currently
supported in the PK module.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Ideally this and other pk functions would work with
MBEDTLS_PSA_CRYPTO_CLIENT (i.e. whether PSA API functions are implemented
locally or via client-server communication). However, at the moment, some
helper functions are missing when MBEDTLS_PSA_CRYPTO_C is disabled, at least
mbedtls_ecc_group_to_psa(). For the time being, don't provide
mbedtls_pk_get_psa_attributes() when MBEDTLS_PSA_CRYPTO_C is disabled. We
can improve later, looking generally at a group of functions to generalize,
not mixed with delivering new APIs.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Putting the id at the of the psa_key_attributes_s structure allows
for a more efficient marshalling of the parameters around a transport
channel which provides separation between a client view and a service
view of the key parameters.
Signed-off-by: Antonio de Angelis <antonio.deangelis@arm.com>
When building without PSA crypto functions, disable
MBEDTLS_PSA_CRYPTO_CLIENT as well as MBEDTLS_PSA_CRYPTO_C. With
just MBEDTLS_PSA_CRYPTO_CLIENT, PSA crypto API functions are supposed to
exist at link time but be provided by a third party.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
In our tests, we run `tests/scripts/doxygen.sh`, which checks that `doxygen`
runs without warnings after `scripts/config.py realfull`. In this
configuration, alternative implementations such as `MBEDTLS_RSA_ALT` are
enabled, which allows the documentation to contain references to the
`MBEDTLS_xxx_ALT` symbol itself. However, this disables context types that
alternative implementations must define in their header, such as
`mbedtls_rsa_context`. See https://github.com/Mbed-TLS/mbedtls/issues/4518
As a partial fix, allow `tests/scripts/doxygen.sh` to see dummy definitions
of the context type. This way, we can use both `#MBEDTLS_RSA_ALT` and
`#mbedtls_rsa_context` cross-references in our documentation. This is not
ideal, because `doxygen.sh` isn't testing for errors in the documentation of
the affected context types, but it's cheap progress.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
This commit makes the code arguably slightly simpler and fixes the build
with clang -Wimplicit-fallthrough. No intended semantic change.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
This isn't officially supported, but it's convenient to build the library
this way for quick testing.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Add code and unit tests for MBEDTLS_PK_ECxxx in
mbedtls_pk_get_psa_attributes().
This commit only supports built-in ECC (MBEDTLS_ECP_C). A subsequent commit
will handle driver-only ECC.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
This avoids a possible gotcha when if the application code reuses an
existing attribute structure.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Upon further consideration we think that a remote attacker close to the
victim might be able to have precise enough timing information to
exploit the side channel as well. Update the Changelog to reflect this.
Signed-off-by: Janos Follath <janos.follath@arm.com>
Any timing variance dependant on the output of this function enables a
Bleichenbacher attack. It is extremely difficult to use safely.
In the Marvin attack paper
(https://people.redhat.com/~hkario/marvin/marvin-attack-paper.pdf) the
author suggests that implementations of PKCS 1.5 decryption that don't
include a countermeasure should be considered inherently dangerous.
They suggest that all libraries implement the same countermeasure, as
implementing different countermeasures across libraries enables the
Bleichenbacher attack as well.
This is extremely fragile and therefore we don't implement it. The use
of PKCS 1.5 in Mbed TLS implements the countermeasures recommended in
the TLS standard (7.4.7.1 of RFC 5246) and is not vulnerable.
Add a warning to PKCS 1.5 decryption to warn users about this.
Signed-off-by: Janos Follath <janos.follath@arm.com>
