Commit graph

29472 commits

Author SHA1 Message Date
Gilles Peskine 6fe8a06f7e New test helper: mbedtls_test_key_consistency_psa_pk
Test that a PK key and a PSA key are consistent, i.e. that they have the
same type (or are a key pair and the corresponding public key) and that
they have the same public key.

Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2024-02-15 17:21:17 +01:00
Gilles Peskine 4da1f01f88 Fix copypasta
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2024-02-15 17:15:29 +01:00
Gilles Peskine d6fc3501c0 Test mbedtls_pk_import_into_psa with different bits
This was only tested with opaque keys. Since the code paths are different
depending on the PK type, we also need to test RSA and ECKEY.

Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2024-02-15 17:15:29 +01:00
Gilles Peskine 68a287dc7a Use named constants FROM_PAIR/FROM_PUBLIC for readability
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2024-02-15 13:30:55 +01:00
Gilles Peskine f50cd59b51 Fix encrypt/decrypt confusion
The values are the same for all supported mechanisms (RSA-based), so no
semantic change.

Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2024-02-15 13:13:26 +01:00
Gilles Peskine cbd2cbb408 Rename identifier for consistency
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2024-02-15 13:12:44 +01:00
Gilles Peskine 465e4ed56b Prioritize SHA2 over MD5 for KNOWN_SUPPORTED_HASH_ALG
This fixes the ability to exercise keys in configurations where MD5 is
supported for direct use, but not inside some accelerated algorithms. This
is the case in `all.sh test_psa_crypto_config_accel_ecc_ecp_light_only` and
some other accelerated-ECC components of `all.sh`, where the driver is built
without MD5 support but built-in MD5 remains enabled.

This is only a hack, not a theoretically correct fix, but a correct fix is
out of scope of my current work.

Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2024-02-12 19:54:53 +01:00
Gilles Peskine 34955677e5 Don't exercise if the algorithm is not supported
Parsing a key and importing it into PSA may result in a policy that
specifies an algorithm that is not included in the build. This happens if
the key type is supported, but not the algorithm, e.g. in a build with
MBEDTLS_ECP_C but not MBEDTLS_ECDSA_C.

Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2024-02-12 19:31:55 +01:00
Gilles Peskine 1d33876d37 Fix some preprocessor guards
Fix the build in some configurations.

Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2024-02-12 19:07:45 +01:00
Gilles Peskine 74860ddec2 Don't define pk_sign_verify in configurations where it's unused
In some configurations (e.g. ECDH but no ECDSA or RSA), the PK module is
useful but cannot perform any signatures. Then modern GCC complains:

```
../source/tests/suites/test_suite_pk.function: In function ‘test_pk_sign_verify’:
../source/tests/suites/test_suite_pk.function:1136:12: error: array subscript 0 is outside array bounds of ‘unsigned char[0]’ [-Werror=array-bounds]
../source/tests/suites/test_suite_pk.function:1094:19: note: while referencing sig’
…
```

This fixes test-ref-configs.pl with a modern GCC (specifically with
config-thread.h).

Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2024-02-12 19:07:45 +01:00
Gilles Peskine 35cb319832 depends.py: set unique configuration names in outcome file
Set unique configuration names in the outcome file. This was lost in the
rewrite from depends-*.pl to depends.py.

Fix #7290

Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2024-02-12 17:32:44 +01:00
Gilles Peskine 157679c0d5 mbedtls_pk_import_into_psa: positive tests with pkparse output
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2024-02-12 17:32:44 +01:00
Gilles Peskine 10e9c412c0 mbedtls_pk_import_into_psa: negative tests for different ECC curve
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2024-02-12 17:32:44 +01:00
Gilles Peskine fc3d866ad2 mbedtls_pk_import_into_psa: implement and test
Implement mbedtls_pk_import_into_psa for all PK types except RSA_ALT.
This covers importing a key pair, importing a public key and importing
the public part of a key pair.

Test mbedtls_pk_import_into_psa() with the output of
mbedtls_pk_get_psa_attributes(). Also unit-test mbedtls_pk_import_into_psa()
on its own to get extra coverage, mostly for negative cases.

Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2024-02-12 17:32:44 +01:00
Gilles Peskine fdb809ef86 exercise_key: fix asymmetric encrypt/decrypt with >2028-bit RSA
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2024-02-12 17:31:03 +01:00
Gilles Peskine 4781bd9773 exercise_key: allow SIGN_MESSAGE/VERIFY_MESSAGE with PSA_ALG_ANY_HASH
There was already code to instantiate the wildcard for sign/verify-hash.
Make that work with sign/verify-message as well.

Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2024-02-12 17:30:27 +01:00
Gilles Peskine 05ee3fbdc0 mbedtls_pk_import_into_psa: documentation
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2024-02-09 19:31:04 +01:00
Gilles Peskine 48b87ebde3 Choose a curve for tests at compile time
This makes it possible to use the curve in test data.

Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2024-02-09 19:31:04 +01:00
Janos Follath 7a28738205
Merge pull request #8636 from paul-elliott-arm/new_test_thread_interface
New test thread interface
2024-02-08 12:35:40 +00:00
Tom Cosgrove 1dbfc8ad3c
Merge pull request #8790 from paul-elliott-arm/fix_ctr_drbg_comment
Fix confusing comment in ctr drbg thread test
2024-02-08 11:11:50 +00:00
Manuel Pégourié-Gonnard b7307630bb
Merge pull request #8703 from valeriosetti/issue7765-guards-in-asn1
Conversion function between raw and DER ECDSA signatures (guards in ASN1)
2024-02-08 08:45:30 +00:00
Manuel Pégourié-Gonnard 7bf1e98f44
Merge pull request #8740 from valeriosetti/issue8647
Move RSA basic key parsing/writing to rsa.c
2024-02-08 08:35:42 +00:00
Tom Cosgrove c8de362202
Merge pull request #8665 from ivq/reduce_static_mem
Reduce many unnecessary static memory consumption
2024-02-07 23:26:27 +00:00
Valerio Setti 1910390b4a psa_util: improve leading zeros check in convert_der_to_raw_single_int()
Signed-off-by: Valerio Setti <valerio.setti@nordicsemi.no>
2024-02-07 16:16:58 +01:00
Paul Elliott bda577bb0b Fix confusing comment in ctr drbg thread test
Make it clearer where the magic number chosen for entropy_len actually
comes from, and why we chose this value.

Signed-off-by: Paul Elliott <paul.elliott@arm.com>
2024-02-07 15:13:46 +00:00
Dave Rodgman 9b272ac1c6
Merge pull request #8794 from daverodgman/pr-guidelines
Remind contributors not to force-push
2024-02-07 15:08:42 +00:00
Valerio Setti ef07fa0fc3 test_suite_psa_crypto_util: add more test for raw->der
Signed-off-by: Valerio Setti <valerio.setti@nordicsemi.no>
2024-02-07 15:16:45 +01:00
Valerio Setti affba30833 psa_util: update documentation for mbedtls_ecdsa_raw_to_der()
Signed-off-by: Valerio Setti <valerio.setti@nordicsemi.no>
2024-02-07 15:03:33 +01:00
Manuel Pégourié-Gonnard b02c0be06a
Merge pull request #8791 from gilles-peskine-arm/psa-legacy-bridges-ecdsa-bits-first
Update ECDSA signature conversion specification
2024-02-07 13:43:29 +00:00
Paul Elliott 292b1dc1e1
Merge pull request #8789 from paul-elliott-arm/fix_tsan_gcc
Stop platform test failures with GCC and TSAN
2024-02-07 11:32:39 +00:00
Dave Rodgman 2a6593bbb6 Slightly soften force-push suggestion
Signed-off-by: Dave Rodgman <dave.rodgman@arm.com>
2024-02-07 11:05:47 +00:00
Dave Rodgman c1a4d1f09a Remove comments about rebasing vs merging; link to longer RTD document
Signed-off-by: Dave Rodgman <dave.rodgman@arm.com>
2024-02-07 11:04:14 +00:00
Dave Rodgman 2840523ae4 Remind contributors not to force-push
Signed-off-by: Dave Rodgman <dave.rodgman@arm.com>
2024-02-07 10:42:41 +00:00
Gilles Peskine 3f557ad59c Wording improvement
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2024-02-07 11:22:16 +01:00
Manuel Pégourié-Gonnard 1d7bc1ecdf
Merge pull request #8717 from valeriosetti/issue8030
PSA FFDH: feature macros for parameters
2024-02-07 10:06:03 +00:00
Dave Rodgman 57a0957938
Merge pull request #8788 from daverodgman/old-gcc-alignment-bug
Change unaligned access method for old gcc
2024-02-07 09:31:45 +00:00
Valerio Setti 447bbce8b4 rsa: remove unnecessary check in priv/pub key parsing
Signed-off-by: Valerio Setti <valerio.setti@nordicsemi.no>
2024-02-07 08:02:03 +01:00
Gilles Peskine 30a303f1a8 ECDSA signature conversion: put bits first
Metadata, then inputs, then outputs.
https://github.com/Mbed-TLS/mbedtls/pull/8703#discussion_r1474697136

Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2024-02-06 19:45:11 +01:00
Paul Elliott e053cb2f12 Stop platform test failures with GCC and TSAN
Signed-off-by: Paul Elliott <paul.elliott@arm.com>
2024-02-06 18:10:43 +00:00
Gilles Peskine 8bdd8cdc4f
Merge pull request #8729 from adeaarm/crypto_struct_client_view
Add a client view of the multipart contexts
2024-02-06 17:29:55 +00:00
Gilles Peskine f45589b492
Merge pull request #8198 from silabs-Kusumit/kdf_incorrect_initial_capacity
KDF incorrect initial capacity
2024-02-06 17:29:43 +00:00
Gilles Peskine 137e0c1a02
Merge pull request #8761 from valeriosetti/issue4681
Re-introduce enum-like checks from CHECK_PARAMS
2024-02-06 17:29:38 +00:00
Gilles Peskine fb7001f15b
Merge pull request #8738 from gilles-peskine-arm/pk_import_into_psa-use_usage
Implement mbedtls_pk_get_psa_attributes
2024-02-06 17:28:54 +00:00
Valerio Setti 1810fd9ac8 add changelog
Signed-off-by: Valerio Setti <valerio.setti@nordicsemi.no>
2024-02-06 17:03:32 +01:00
Valerio Setti bb76f80218 pk_wrap: use proper raw buffer length in ecdsa_sign_psa()
Signed-off-by: Valerio Setti <valerio.setti@nordicsemi.no>
2024-02-06 16:57:23 +01:00
Valerio Setti cf81f69977 psa_util: smarter raw length check in mbedtls_ecdsa_raw_to_der()
Signed-off-by: Valerio Setti <valerio.setti@nordicsemi.no>
2024-02-06 16:57:12 +01:00
Dave Rodgman 91d5fde944
Merge pull request #8745 from adeaarm/trail_key_id_field
Put the id field at the end of the psa_key_attributes_s structure
2024-02-06 15:55:56 +00:00
Valerio Setti 6269f3baf4 Revert "psa_util: allow larger raw buffers in mbedtls_ecdsa_raw_to_der()"
This reverts commit d4fc5d9d1c.

Signed-off-by: Valerio Setti <valerio.setti@nordicsemi.no>
2024-02-06 16:55:18 +01:00
Valerio Setti 2b6a7b37f4 suite_psa_crypto_util: use 521 bits data and bit-size instead of 528
Signed-off-by: Valerio Setti <valerio.setti@nordicsemi.no>
2024-02-06 16:21:44 +01:00
Dave Rodgman e093281a8b Pacify check-names
Signed-off-by: Dave Rodgman <dave.rodgman@arm.com>
2024-02-06 15:00:58 +00:00