This is an external function, so in the absence of link-time
optimisation (LTO) the compiler can't know anything about it and has to
call it the number of times it's called in the source code.
This only matters for pk_ec, but change pk_rsa as well for the sake of
uniformity.
Signed-off-by: Manuel Pégourié-Gonnard <manuel.pegourie-gonnard@arm.com>
Trusting the caller to perform the appropriate check is both risky, and
a bit user-unfriendly. Returning NULL on error seems both safer
(dereferencing a NULL pointer is more likely to result in a clean crash,
while mis-casting a pointer might have deeper, less predictable
consequences) and friendlier (the caller can just check the return
value for NULL, which is a common idiom).
Only add that as an additional way of using the function, for the sake
of backwards compatibility. Calls where we know the type of the context
for sure (for example because we just set it up) were legal and safe, so
they should remain legal without checking the result for NULL, which
would be redundant.
Signed-off-by: Manuel Pégourié-Gonnard <manuel.pegourie-gonnard@arm.com>
The previous wording "ensure it holds an XXX" context did not mean
anything without looking at the source.
Looking at the source, the criterion is:
- for mbedtls_pk_rsa(), that the info structure uses rsa_alloc_wrap;
- for mbedtls_pk_ec(), that it uses eckey_alloc_wrap or
ecdsa_alloc_wrap, since mbedtls_ecdsa_context is a typedef for
mbedtls_ecp_keypair. (Note that our test code uses mbedtls_pk_ec() on
contexts of type MBEDTLS_PK_ECDSA.)
Signed-off-by: Manuel Pégourié-Gonnard <manuel.pegourie-gonnard@arm.com>
Promise that we will try to keep backward compatibility with basic driver
usage, but not with more experimental aspects.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
This was intended as experimental, and we've been saying for a long time
that it's superseded by the "unified driver interface", but we hadn't
documented that inside the Mbed TLS source code. So announce it as
deprecated.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Certain numerical values are written to the key store. Changing those
numerical values would break the backward compatibility of stored keys. Add
a note to the affected types. Add comments near the definitions of affected
values.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Move the definition of the accessor so that it is not defined
within the MBEDTLS_X509_CRT_WRITE_C guards. Thus remove the
dependency from the test and test cases.
Signed-off-by: Thomas Daubney <thomas.daubney@arm.com>
Remaining hits seem to be hex data, certificates,
and other miscellaneous exceptions.
List generated by running codespell -w -L
keypair,Keypair,KeyPair,keyPair,ciph,nd
Signed-off-by: Andrzej Kurek <andrzej.kurek@arm.com>
The accessor macros for key_id and owner_id in the mbedtls_svc_key_id_t
need to have the MBEDTLS_PRIVATE() specifier as these fields are private
Signed-off-by: Antonio de Angelis <antonio.deangelis@arm.com>
This commit fixes#1992: The documentation of mbedtls_x509_crt_profile
previously stated that the bitfield `allowed_pks` defined which signature
algorithms shall be allowed in CRT chains. In actual fact, however,
the field also applies to guard the public key of the end entity
certificate.
This commit changes the documentation to state that `allowed_pks`
applies to the public keys of all CRTs in the provided chain.
Signed-off-by: Manuel Pégourié-Gonnard <manuel.pegourie-gonnard@arm.com>
Remove useless hrr code
Share validate_cipher_suit between client and server
Fix test failure when tls13 only in server side
Change-Id: I5d6a7932bd8448ebf542bc86cdcab8862bc28e9b
Signed-off-by: XiaokangQian <xiaokang.qian@arm.com>
Refine named_group parsing
Refine cipher_suites parsing
Remove hrr related part
Share code between client and server side
Some code style changes
Change-Id: Ia9ffd5ef9c0b64325f633241e0ea1669049fe33a
Signed-off-by: XiaokangQian <xiaokang.qian@arm.com>
Remove cookie support from server side
Change code to align with coding styles
Re-order functions of client_hello
Change-Id: If31509ece402f8276e6cac37f261e0b166d05e18
Signed-off-by: XiaokangQian <xiaokang.qian@arm.com>
The cipher module implements XTS, and the PSA API specifies XTS, but the PSA
implementation does not support XTS. It requires double-size keys, which
psa_crypto does not currently support.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Use mbedtls_ssl_protocol_version in public structs, even when doing
so results in a binary-incompatible change to the public structure
(PR feedback from @ronald-cron-arm)
Signed-off-by: Glenn Strauss <gstrauss@gluelogic.com>
Store the TLS version in tls_version instead of major, minor version num
Note: existing application use which accesses the struct member
(using MBEDTLS_PRIVATE) is not compatible, as the struct is now smaller.
Reduce size of mbedtls_ssl_ciphersuite_t
members are defined using integral types instead of enums in
order to pack structure and reduce memory usage by internal
ciphersuite_definitions[]
Signed-off-by: Glenn Strauss <gstrauss@gluelogic.com>
Store the TLS version in tls_version instead of major, minor version num
Note: existing application use which accesses the struct member
(using MBEDTLS_PRIVATE) is not compatible on little-endian platforms,
but is compatible on big-endian platforms. For systems supporting
only TLSv1.2, the underlying values are the same (=> 3).
New setter functions are more type-safe,
taking argument as enum mbedtls_ssl_protocol_version:
mbedtls_ssl_conf_max_tls_version()
mbedtls_ssl_conf_min_tls_version()
Signed-off-by: Glenn Strauss <gstrauss@gluelogic.com>
Store the TLS version instead of minor version number in tls_version.
Note: struct member size changed from unsigned char to uint16_t
Due to standard structure padding, the structure size does not change
unless alignment is 1-byte (instead of 2-byte or more)
Note: existing application use which accesses the struct member
(using MBEDTLS_PRIVATE) is compatible on little-endian platforms,
but not compatible on big-endian platforms. The enum values for
the lower byte of MBEDTLS_SSL_VERSION_TLS1_2 and of
MBEDTLS_SSL_VERSION_TLS1_3 matches MBEDTLS_SSL_MINOR_VERSION_3 and
MBEDTLS_SSL_MINOR_VERSION_4, respectively.
Note: care has been taken to preserve serialized session format,
which uses only the lower byte of the TLS version.
Signed-off-by: Glenn Strauss <gstrauss@gluelogic.com>
This remark is intended for maintainers, not for users. It should not have
been in the Doxygen typeset part.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
With Doxygen 1.8.11 (as on Ubuntu 16.04), `#include` doesn't protect the
hash character enough, and Doxygen tries to link to something called
include. (Doxygen 1.8.17 doesn't have this problem.)
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Include this new section in the "full for documentation" (`realfull`)
configuration, so that these options are documented in the official
documentation build (`scripts/apidoc_full.sh`).
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
The timing module might include time.h on its own when on
a suitable platform, even if MBEDTLS_HAVE_TIME is disabled.
Co-authored-by: Tom Cosgrove <tom.cosgrove@arm.com>
Signed-off-by: Andrzej Kurek <andrzej.kurek@arm.com>
When MBEDTLS_PSA_CRYPTO_CONFIG is enabled, support an alternative file to
include instead of "psa/crypto_config.h", and an additional file to include
after it. This follows the model of the existing MBEDTLS_{,USER_}CONFIG_FILE.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Rename ssl_cli.c and ssl_srv.c to reflect the fact
that they are TLS 1.2 specific now. Align there new
names with the TLS 1.3 ones.
Signed-off-by: Ronald Cron <ronald.cron@arm.com>
Provide an additional pair of #defines, MBEDTLS_SHA512_USE_A64_CRYPTO_IF_PRESENT
and MBEDTLS_SHA512_USE_A64_CRYPTO_ONLY. At most one of them may be
specified. If used, it is necessary to compile with -march=armv8.2-a+sha3.
The MBEDTLS_SHA512_PROCESS_ALT and MBEDTLS_SHA512_ALT mechanisms
continue to work, and are mutually exclusive with SHA512_USE_A64_CRYPTO.
There should be minimal code size impact if no A64_CRYPTO option is set.
The SHA-512 implementation was originally written by Simon Tatham for PuTTY,
under the MIT licence; dual-licensed as Apache 2 with his kind permission.
Signed-off-by: Tom Cosgrove <tom.cosgrove@arm.com>
Add function to query if SSL handshake is over or not, in order to
determine when to stop calling mbedtls_ssl_handshake_step among other
things. Document function, and add warnings that the previous method of
ascertaining if handshake was over is now deprecated, and may break in
future releases.
Signed-off-by: Paul Elliott <paul.elliott@arm.com>
- parameter name in function description
- test_suite_ecp.data: add new line at the end of file
Signed-off-by: Przemek Stekiel <przemyslaw.stekiel@mobica.com>
CCM*-no-tag is currently available whenever CCM is, so declare
PSA_WANT_ALG_CCM_STAR_NO_TAG whenever PSA_WANT_ALG_CCM is declared and vice
versa.
Fix dependencies of test cases that use PSA_ALG_CCM_STAR_NO_TAG: some were
using PSA_WANT_ALG_CCM and some had altogether wrong dependencies.
This commit does not touch library code. There is still no provision for
providing CCM support without CCM*-no-tag or vice versa.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
MBEDTLS_HAVE_TIME is documented as: "System has time.h and time()."
If that is not defined, do not attempt to include time.h.
A particular problem is platform-time.h, which should only be included if
MBEDTLS_HAVE_TIME is defined, which makes everything messier. Maybe it
should be refactored to have the check inside the header.
Signed-off-by: Daniel Axtens <dja@axtens.net>
Extend mbedtls_ssl_set_hs_own_cert() to reset handshake cert list
if cert provided is null. Previously, mbedtls_ssl_set_hs_own_cert()
only provided a way to append to the handshake certificate list,
without providing a way to replace the handshake certificate list.
Signed-off-by: Glenn Strauss <gstrauss@gluelogic.com>
Perform the following optimizations:
- fix used flags for conditional compilation
- remove redundant N variable
- move loop used to generate valid k value to helper function
- fix initial value of status
- fix comments
Signed-off-by: Przemyslaw Stekiel <przemyslaw.stekiel@mobica.com>
Mention that TLS 1.3 is supported, in addition to (D)TLS 1.2.
Improve and clarify the documentation. In particular, emphasise that the
minor version numbers are the internal numbers which are off by one from the
human numbers.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Same intended semantics, no casts.
Limitation: this doesn't work on architectures where
sizeof(uintptr_t) < sizeof(void*), which is somewhat weird but possible if
pointers contain redundant information.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
The user data is typically a pointer to a data structure or a handle which
may no longer be valid after the session is restored. If the user data needs
to be preserved, let the application do it. This way, it is a conscious
decision for the application to save/restore either the pointer/handle
itself or the object it refers to.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
In structure types that are passed to user callbacks, add a field that the
library won't ever care about. The application can use this field to either
identify an instance of the structure with a handle, or store a pointer to
extra data.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Provide an additional pair of #defines, MBEDTLS_SHA256_USE_A64_CRYPTO_IF_PRESENT
and MBEDTLS_SHA256_USE_A64_CRYPTO_ONLY. At most one of them may be
specified. If used, it is necessary to compile with -march=armv8-a+crypto.
The MBEDTLS_SHA256_PROCESS_ALT and MBEDTLS_SHA256_ALT mechanisms
continue to work, and are mutually exclusive with A64_CRYPTO.
There should be minimal code size impact if no A64_CRYPTO option is set.
Signed-off-by: Tom Cosgrove <tom.cosgrove@arm.com>
These fields need to be enabled for 1.3 even if MBEDTLS_USE_PSA_CRYPTO isn't (1.3 should always use PSA).
Signed-off-by: Przemyslaw Stekiel <przemyslaw.stekiel@mobica.com>
Initially this function was doing something because the output format of
psa_export_public() didn't match the ECPoint format that TLS wants.
Then it became a no-op then the output format of psa_export_public()
changed, but it made sense to still keep the function in case the format
changed again. Now that the PSA Crypto API has reached 1.0 status, this
is unlikely to happen, so the no-op function is no longer useful.
Removing it de-clutters the code a bit; while at it we can remove a
temporary stack buffer (that was up to 133 bytes).
It's OK to remove this function even if it was declared in a public
header, as there's a warning at the top of the file saying it's not part
of the public API.
Signed-off-by: Manuel Pégourié-Gonnard <manuel.pegourie-gonnard@arm.com>
Relying on a PSA_VENDOR macro is not ideal, since the standard doesn't
guarantee this macro exists, but OTOH relying on
MBEDTLS_ECP_DP_xxx_ENABLED was even less ideal, so I believe this is
still an improvement.
Signed-off-by: Manuel Pégourié-Gonnard <manuel.pegourie-gonnard@arm.com>
We had ECC then PK then ECC, move PK to the end, now all ECC things are
together. (The comments suggest that was the intention all along.)
Signed-off-by: Manuel Pégourié-Gonnard <manuel.pegourie-gonnard@arm.com>
PKCS5 depends on MD, but is missing a config check resulting in
obscure errors on invalid configurations.
Signed-off-by: Sebastian Bøe <sebastian.boe@nordicsemi.no>
At the end of the benchmark program, heap stats are printed, and these
stats will be wrong if we reset counters in the middle.
Also remove the function to reset counters, in order to encourage other
programs to behave correctly as well.
Signed-off-by: Manuel Pégourié-Gonnard <manuel.pegourie-gonnard@arm.com>
Fix library references, tests and programs.
Testing is performed in the already present all.sh test.
Signed-off-by: Andrzej Kurek <andrzej.kurek@arm.com>
Declare mbedtls_md functions as MBEDTLS_CHECK_RETURN_TYPICAL, meaning that
their return values should be checked.
Do check the return values in our code. We were already doing that
everywhere for hash calculations, but not for HMAC calculations.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
As we have now a minimal viable implementation of TLS 1.3,
let's remove EXPERIMENTAL from the config option enabling
it.
Signed-off-by: Ronald Cron <ronald.cron@arm.com>
Ensure that the documentation of fields affected by
"mbedtls_ssl_config: Replace bit-fields by separate bytes"
conveys information that may have been lost by removing the exact size of
the type. Extend the preexisting pattern "do this?" for formerly 1-bit
boolean fields. Indicate the possible values for non-boolean fields.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Change the wording of the documentation for some CMAC functions,
as the existing wording, while technically correct, can be
easy to misunderstand. The reworded docs explain the flow of
a CMAC computation a little more fully.
Signed-off-by: David Horstmann <david.horstmann@arm.com>
Use PSA_BUILTIN macros instead of the Mbed TLS ones
as in the hash operation contexts the context for a
given hash is needed only if the support for it
through PSA is enabled.
Signed-off-by: Ronald Cron <ronald.cron@arm.com>
This commit removes the test_psa_crypto_config_basic
all.sh component that can no longer work without
adapting it to the separately compiled test driver
library. This component is replaced by several
components in the following commits to test various
type of acceleration independently.
Signed-off-by: Ronald Cron <ronald.cron@arm.com>
The PSA max hash size has to be 64 if SHA512 or
SHA384 is supported by the library or an
accelerator, not just in case of the library.
Signed-off-by: Ronald Cron <ronald.cron@arm.com>
static function mbedtls_set_key_owner() is declared in psa/crypto.h
and defined in psa/crypto_struct.h with different parameter name for
the mbedtls_key_owner_id_t parameter and that may trigger errors
from static code analysis tool as cppcheck.
Signed-off-by: Antonio de Angelis <antonio.deangelis@arm.com>
This slightly increases the RAM consumption per context, but saves code
size on architectures with an instruction for direct byte access (which is
most of them).
Although this is technically an API break, in practice, a realistic
application won't break: it would have had to bypass API functions and rely
on the field size (e.g. relying on -1 == 1 in a 1-bit field).
Results (arm-none-eabi-gcc 7.3.1, build_arm_none_eabi_gcc_m0plus build):
library/ssl_cli.o: 19543 -> 19559 (diff: -16)
library/ssl_msg.o: 24726 -> 24690 (diff: 36)
library/ssl_srv.o: 20462 -> 20418 (diff: 44)
library/ssl_tls.o: 20707 -> 20555 (diff: 152)
library/ssl_tls13_client.o: 7252 -> 7244 (diff: 8)
library/ssl_tls13_generic.o: 4705 -> 4693 (diff: 12)
Results (same architecture, config-suite-b.h + MBEDTLS_ECDH_LEGACY_CONTEXT +
MBEDTLS_ECP_RESTARTABLE):
library/ssl_cli.o: 2876 -> 2864 (diff: 12)
library/ssl_msg.o: 3068 -> 3080 (diff: -12)
library/ssl_srv.o: 3372 -> 3340 (diff: 32)
library/ssl_tls.o: 6658 -> 6566 (diff: 92)
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Move small fields first so that more fields can be within the Arm Thumb
128-element direct access window.
Keep the int section after the pointer section: moving int fields first cost
a few bytes on the reference baremetal-m0plus build.
The ordering in this commit is not based on field access frequency.
Results (arm-none-eabi-gcc 7.3.1, build_arm_none_eabi_gcc_m0plus build):
library/ssl_cli.o: 19687 -> 19543 (diff: 144)
library/ssl_msg.o: 24834 -> 24726 (diff: 108)
library/ssl_srv.o: 20562 -> 20462 (diff: 100)
library/ssl_tls.o: 20907 -> 20707 (diff: 200)
library/ssl_tls13_client.o: 7272 -> 7252 (diff: 20)
library/ssl_tls13_generic.o: 4721 -> 4705 (diff: 16)
Results (same architecture, config-suite-b.h + MBEDTLS_ECDH_LEGACY_CONTEXT +
MBEDTLS_ECP_RESTARTABLE):
library/ssl_cli.o: 2936 -> 2876 (diff: 60)
library/ssl_msg.o: 3080 -> 3068 (diff: 12)
library/ssl_srv.o: 3400 -> 3372 (diff: 28)
library/ssl_tls.o: 6730 -> 6658 (diff: 72)
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Move fields around to have fewer accesses outside the 128-element Thumb
direct access window.
In psa_hkdf_key_derivation_t, move the large fields (output_block, prk,
hmac) after the state bit-fields. Experimentally, it's slightly better
to put hmac last.
Other operations structures don't go outside the window, at least when not
considering nested structures.
Results (arm-none-eabi-gcc 7.3.1, build_arm_none_eabi_gcc_m0plus build):
library/psa_crypto.o: 16510 -> 16434 (diff: 76)
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Several files among include/psa/crypto_*.h are not meant to be included
directly, and are not guaranteed to be valid if included directly. This
makes it harder to perform some static analyses. So make these files more
self-contained so that at least, if included on their own, there is no
missing macro or type definition (excluding the deliberate use of forward
declarations of structs and unions).
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Previously passing a NULL or zero length password into either
mbedtls_pkcs12_pbe() or mbedtls_pkcs12_derive() could cause an infinate
loop, and it was also possible to pass a NULL password, with a non-zero
length, which would cause memory corruption.
I have fixed these errors, and improved the documentation to reflect the
changes and further explain what is expected of the inputs.
Signed-off-by: Paul Elliott <paul.elliott@arm.com>
Remove client certificate verify in tests.
Change the layout of structure to fix abi_api check issues.
Add comments of Finished.
Align with the coding styles.
Signed-off-by: XiaokangQian <xiaokang.qian@arm.com>
Fix the variable not inialized issue, remove the client
certificate related code, remove early data related code.
Signed-off-by: XiaokangQian <xiaokang.qian@arm.com>
The current definition of PSA_ALG_IS_HASH_AND_SIGN includes
PSA_ALG_RSA_PKCS1V15_SIGN_RAW and PSA_ALG_ECDSA_ANY, which don't strictly
follow the hash-and-sign paradigm: the algorithm does not encode a hash
algorithm that is applied prior to the signature step. The definition in
fact encompasses what can be used with psa_sign_hash/psa_verify_hash, so
it's the correct definition for PSA_ALG_IS_SIGN_HASH. Therefore this commit
moves definition of PSA_ALG_IS_HASH_AND_SIGN to PSA_ALG_IS_SIGN_HASH, and
replace the definition of PSA_ALG_IS_HASH_AND_SIGN by a correct one (based
on PSA_ALG_IS_SIGN_HASH, excluding the algorithms where the pre-signature
step isn't to apply the hash encoded in the algorithm).
In the definition of PSA_ALG_SIGN_GET_HASH, keep the condition for a nonzero
output to be PSA_ALG_IS_HASH_AND_SIGN.
Everywhere else in the code base (definition of PSA_ALG_IS_SIGN_MESSAGE, and
every use of PSA_ALG_IS_HASH_AND_SIGN outside of crypto_values.h), we meant
PSA_ALG_IS_SIGN_HASH where we wrote PSA_ALG_IS_HASH_AND_SIGN, so do a
global replacement.
```
git grep -l IS_HASH_AND_SIGN ':!include/psa/crypto_values.h' | xargs perl -i -pe 's/ALG_IS_HASH_AND_SIGN/ALG_IS_SIGN_HASH/g'
```
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
mbedtls_ssl_conf_groups allows supported groups for key
sharing to be configured via their IANA NamedGroup ID.
This is added in anticipation of PQC and Hybrid key
sharing algorithms being integrated into Mbed TLS.
mbedtls_ssl_conf_curves is deprecated in favor of
mbedtls_ssl_conf_groups. handshake_init has been
modified to translate and copy curves configured
via conf_curves into a heap allocatied array of
NamedGroup IDs. This allows the refactoring of code
interacting with conf_curve related variables (such
as curve_list) to use NamedGroup IDs while retaining
the deprecated API.
Signed-off-by: Brett Warren <brett.warren@arm.com>
Since there are no longer any alternative
MBEDTLS_DEPRECATED definitions in the codebase,
MBEDTLS_DEPRECATED can now be exported without breaking
anything.
Signed-off-by: Brett Warren <brett.warren@arm.com>
TLS 1.3: ServerHello: add utils functions used by ServerHello
Regarding the merge job, there was only one of the failure we currently encounter on almost all PR (Session resume using tickets, DTLS: openssl client test case see #5012) thus we can consider that this PR passed CI.
This is a variant of PSA_ALG_RSA_PSS which currently has exactly the same
behavior, but is intended to have a different behavior when verifying
signatures.
In a subsequent commit, PSA_ALG_RSA_PSS will change to requiring the salt
length to be what it would produce when signing, as is currently documented,
whereas PSA_ALG_RSA_PSS_ANY_SALT will retain the current behavior of
allowing any salt length (including 0).
Changes in this commit:
* New algorithm constructor PSA_ALG_RSA_PSS_ANY_SALT.
* New predicates PSA_ALG_IS_RSA_PSS_STANDARD_SALT (corresponding to
PSA_ALG_RSA_PSS) and PSA_ALG_IS_RSA_PSS_ANY_SALT (corresponding to
PSA_ALG_RSA_PSS_ANY_SALT).
* Support for the new predicates in macro_collector.py (needed for
generate_psa_constant_names).
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
The requirement of minimum 15 bytes for output buffer in
psa_aead_finish() and psa_aead_verify() does not apply
to the built-in implementation of the GCM.
Alternative implementations are expected to verify the
length of the provided output buffers and to return
the MBEDTLS_ERR_GCM_BUFFER_TOO_SMALL in case the
buffer length is too small.
Signed-off-by: Mateusz Starzyk <mateusz.starzyk@mobica.com>
