yuzu-emu/mbedtls
2
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity
mbedtls/tests/suites/test_suite_psa_crypto_metadata.function

740 lines
32 KiB
Text
Raw Normal View History

Create infrastructure for metadata validation unit tests
2018-08-21 18:20:20 +02:00
/* BEGIN_HEADER */
/* Test macros that provide metadata about algorithms and key types.
* This test suite only contains tests that don't require executing
* code. Other test suites validate macros that require creating a key
* and using it. */
#if defined(MBEDTLS_PSA_CRYPTO_SPM)
#include "spm/psa_defs.h"
#endif
#include "psa/crypto.h"
Make psa_get_mac_output_length testable and test it Signed-off-by: Steven Cooreman <steven.cooreman@silabs.com>
2021-03-02 21:31:17 +01:00
#include "psa_crypto_invasive.h"
Create infrastructure for metadata validation unit tests
2018-08-21 18:20:20 +02:00
Algorithm validation tests For all algorithms, validate feature test macros (PSA_ALG_IS_xxx). For hash algorithms, validate the exact hash size, and validate xxx_GET_HASH macros on dependent algorithms. For MAC algorithms, validate the MAC size. For AEAD algorithms, validate the tag size. There is a separate test case for each HMAC algorithm, which is necessary because each has its own MAC size. For other hash-dependent algorithms, there is no interesting variation to test here, so only one hash gets tested.
2018-08-22 18:36:36 +02:00
/* Flags for algorithm classification macros. There is a flag for every
* algorithm classification macro PSA_ALG_IS_xxx except for the
* category test macros, which are hard-coded in each
* category-specific function. The name of the flag is the name of the
* classification macro without the PSA_ prefix. */
#define ALG_IS_VENDOR_DEFINED ( 1u << 0 )
#define ALG_IS_HMAC ( 1u << 1 )
#define ALG_IS_BLOCK_CIPHER_MAC ( 1u << 2 )
#define ALG_IS_STREAM_CIPHER ( 1u << 3 )
#define ALG_IS_RSA_PKCS1V15_SIGN ( 1u << 4 )
#define ALG_IS_RSA_PSS ( 1u << 5 )
Update metadata tests with the new IS_ALG_RSA_PSS_xxx_SALT predicates Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-10-04 18:14:59 +02:00
#define ALG_IS_RSA_PSS_ANY_SALT ( 1u << 6 )
#define ALG_IS_RSA_PSS_STANDARD_SALT ( 1u << 7 )
#define ALG_IS_DSA ( 1u << 8 )
#define ALG_DSA_IS_DETERMINISTIC ( 1u << 9 )
#define ALG_IS_DETERMINISTIC_DSA ( 1u << 10 )
#define ALG_IS_RANDOMIZED_DSA ( 1u << 11 )
#define ALG_IS_ECDSA ( 1u << 12 )
#define ALG_ECDSA_IS_DETERMINISTIC ( 1u << 13 )
#define ALG_IS_DETERMINISTIC_ECDSA ( 1u << 14 )
#define ALG_IS_RANDOMIZED_ECDSA ( 1u << 15 )
#define ALG_IS_HASH_EDDSA ( 1u << 16 )
Add PSA_ALG_IS_HASH_AND_SIGN to the metadata tests The status of signature wildcards with respect to PSA_ALG_IS_HASH_AND_SIGN is unclear in the specification. A wildcard is usually instantiated with a specific hash, making the implementation hash-and-sign, but it could also be instantiated with a non-hash-and-sign algorithm. For the time being, go with what's currently implemented, which is that they are considered hash-and-sign. Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-09-22 18:12:31 +02:00
#define ALG_IS_SIGN_HASH ( 1u << 17 )
#define ALG_IS_HASH_AND_SIGN ( 1u << 18 )
#define ALG_IS_RSA_OAEP ( 1u << 19 )
#define ALG_IS_HKDF ( 1u << 20 )
#define ALG_IS_FFDH ( 1u << 21 )
#define ALG_IS_ECDH ( 1u << 22 )
#define ALG_IS_WILDCARD ( 1u << 23 )
#define ALG_IS_RAW_KEY_AGREEMENT ( 1u << 24 )
#define ALG_IS_AEAD_ON_BLOCK_CIPHER ( 1u << 25 )
#define ALG_IS_TLS12_PRF ( 1u << 26 )
#define ALG_IS_TLS12_PSK_TO_MS ( 1u << 27 )
Ensure that all flags are actually tested At least twice, we added a classification flag but forgot to test it in the relevant test functions. Add some protection so that this doesn't happen again. In each classification category, put a macro xxx_FLAG_MASK_PLUS_ONE at the end. In the corresponding test function, keep track of the flags that are tested, and check that their mask is xxx_FLAG_MASK_PLUS_ONE - 1 which is all the bits of the previous flags set. Now, if we add a flag without testing it, the test TEST_EQUAL( classification_flags_tested, xxx_FLAG_MASK_PLUS_ONE - 1 ) will fail. It will also fail if we make the set of flag numbers non-consecutive, which is ok. This reveals that three algorithm flags had been added but not tested (in two separate occasions). Also, one key type flag that is no longer used by the library was still defined but not tested, which is not a test gap but is inconsistent. It's for DSA, which is relevant to the PSA encoding even if Mbed TLS doesn't implement it, so keep the flag and do test it. Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-11-03 14:18:08 +01:00
#define ALG_FLAG_MASK_PLUS_ONE ( 1u << 28 ) /* must be last! */
Algorithm validation tests For all algorithms, validate feature test macros (PSA_ALG_IS_xxx). For hash algorithms, validate the exact hash size, and validate xxx_GET_HASH macros on dependent algorithms. For MAC algorithms, validate the MAC size. For AEAD algorithms, validate the tag size. There is a separate test case for each HMAC algorithm, which is necessary because each has its own MAC size. For other hash-dependent algorithms, there is no interesting variation to test here, so only one hash gets tested.
2018-08-22 18:36:36 +02:00
Key type validation tests For all key types, validate feature test macros (PSA_KEY_TYPE_IS_xxx). For asymmetric keys (public key or key pair), validate the corresponding public/pair type. For ECC keys, validate GET_CURVE.
2018-08-22 18:43:09 +02:00
/* Flags for key type classification macros. There is a flag for every
* key type classification macro PSA_KEY_TYPE_IS_xxx except for some that
* are tested as derived from other macros. The name of the flag is
* the name of the classification macro without the PSA_ prefix. */
#define KEY_TYPE_IS_VENDOR_DEFINED ( 1u << 0 )
#define KEY_TYPE_IS_UNSTRUCTURED ( 1u << 1 )
#define KEY_TYPE_IS_PUBLIC_KEY ( 1u << 2 )
Add PSA_ALG_IS_HASH_AND_SIGN to the metadata tests The status of signature wildcards with respect to PSA_ALG_IS_HASH_AND_SIGN is unclear in the specification. A wildcard is usually instantiated with a specific hash, making the implementation hash-and-sign, but it could also be instantiated with a non-hash-and-sign algorithm. For the time being, go with what's currently implemented, which is that they are considered hash-and-sign. Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-09-22 18:12:31 +02:00
#define KEY_TYPE_IS_KEY_PAIR ( 1u << 3 )
Key type validation tests For all key types, validate feature test macros (PSA_KEY_TYPE_IS_xxx). For asymmetric keys (public key or key pair), validate the corresponding public/pair type. For ECC keys, validate GET_CURVE.
2018-08-22 18:43:09 +02:00
#define KEY_TYPE_IS_RSA ( 1u << 4 )
#define KEY_TYPE_IS_DSA ( 1u << 5 )
#define KEY_TYPE_IS_ECC ( 1u << 6 )
Parametrize Diffie-Hellman keys by a group identifier Parametrize finite-field Diffie-Hellman key types with a DH group identifier, in the same way elliptic curve keys are parametrized with an EC curve identifier. Define the DH groups from the TLS registry (these are the groups from RFC 7919). Replicate the macro definitions and the metadata tests from elliptic curve identifiers to DH group identifiers. Define PSA_DH_GROUP_CUSTOM as an implementation-specific extension for which domain parameters are used to specify the group.
2019-05-16 12:55:35 +02:00
#define KEY_TYPE_IS_DH ( 1u << 7 )
Ensure that all flags are actually tested At least twice, we added a classification flag but forgot to test it in the relevant test functions. Add some protection so that this doesn't happen again. In each classification category, put a macro xxx_FLAG_MASK_PLUS_ONE at the end. In the corresponding test function, keep track of the flags that are tested, and check that their mask is xxx_FLAG_MASK_PLUS_ONE - 1 which is all the bits of the previous flags set. Now, if we add a flag without testing it, the test TEST_EQUAL( classification_flags_tested, xxx_FLAG_MASK_PLUS_ONE - 1 ) will fail. It will also fail if we make the set of flag numbers non-consecutive, which is ok. This reveals that three algorithm flags had been added but not tested (in two separate occasions). Also, one key type flag that is no longer used by the library was still defined but not tested, which is not a test gap but is inconsistent. It's for DSA, which is relevant to the PSA encoding even if Mbed TLS doesn't implement it, so keep the flag and do test it. Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-11-03 14:18:08 +01:00
#define KEY_TYPE_FLAG_MASK_PLUS_ONE ( 1u << 8 ) /* must be last! */
Key type validation tests For all key types, validate feature test macros (PSA_KEY_TYPE_IS_xxx). For asymmetric keys (public key or key pair), validate the corresponding public/pair type. For ECC keys, validate GET_CURVE.
2018-08-22 18:43:09 +02:00
Add lifetime metadata tests Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-04-21 20:03:53 +02:00
/* Flags for lifetime classification macros. There is a flag for every
* lifetime classification macro PSA_KEY_LIFETIME_IS_xxx. The name of the
* flag is the name of the classification macro without the PSA_ prefix. */
#define KEY_LIFETIME_IS_VOLATILE ( 1u << 0 )
New macro PSA_KEY_LIFETIME_IS_READ_ONLY Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-04-21 20:05:59 +02:00
#define KEY_LIFETIME_IS_READ_ONLY ( 1u << 1 )
Ensure that all flags are actually tested At least twice, we added a classification flag but forgot to test it in the relevant test functions. Add some protection so that this doesn't happen again. In each classification category, put a macro xxx_FLAG_MASK_PLUS_ONE at the end. In the corresponding test function, keep track of the flags that are tested, and check that their mask is xxx_FLAG_MASK_PLUS_ONE - 1 which is all the bits of the previous flags set. Now, if we add a flag without testing it, the test TEST_EQUAL( classification_flags_tested, xxx_FLAG_MASK_PLUS_ONE - 1 ) will fail. It will also fail if we make the set of flag numbers non-consecutive, which is ok. This reveals that three algorithm flags had been added but not tested (in two separate occasions). Also, one key type flag that is no longer used by the library was still defined but not tested, which is not a test gap but is inconsistent. It's for DSA, which is relevant to the PSA encoding even if Mbed TLS doesn't implement it, so keep the flag and do test it. Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-11-03 14:18:08 +01:00
#define KEY_LIFETIME_FLAG_MASK_PLUS_ONE ( 1u << 2 ) /* must be last! */
Add lifetime metadata tests Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-04-21 20:03:53 +02:00
Ensure that all flags are actually tested At least twice, we added a classification flag but forgot to test it in the relevant test functions. Add some protection so that this doesn't happen again. In each classification category, put a macro xxx_FLAG_MASK_PLUS_ONE at the end. In the corresponding test function, keep track of the flags that are tested, and check that their mask is xxx_FLAG_MASK_PLUS_ONE - 1 which is all the bits of the previous flags set. Now, if we add a flag without testing it, the test TEST_EQUAL( classification_flags_tested, xxx_FLAG_MASK_PLUS_ONE - 1 ) will fail. It will also fail if we make the set of flag numbers non-consecutive, which is ok. This reveals that three algorithm flags had been added but not tested (in two separate occasions). Also, one key type flag that is no longer used by the library was still defined but not tested, which is not a test gap but is inconsistent. It's for DSA, which is relevant to the PSA encoding even if Mbed TLS doesn't implement it, so keep the flag and do test it. Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-11-03 14:18:08 +01:00
/* Check that in the value of flags, the bit flag (which should be a macro
* expanding to a number of the form 1 << k) is set if and only if
* PSA_##flag(alg) is true.
*
* Only perform this check if cond is true. Typically cond is 1, but it can
* be different if the value of the flag bit is only specified under specific
* conditions.
*
* Unconditionally mask flag into the ambient variable
* classification_flags_tested.
*/
#define TEST_CLASSIFICATION_MACRO( cond, flag, alg, flags ) \
do \
{ \
if( cond ) \
{ \
if( ( flags ) & ( flag ) ) \
TEST_ASSERT( PSA_##flag( alg ) ); \
else \
TEST_ASSERT( ! PSA_##flag( alg ) ); \
} \
classification_flags_tested |= ( flag ); \
} \
More informative message on invalid classification flags Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-04-21 20:04:24 +02:00
while( 0 )
Algorithm validation tests For all algorithms, validate feature test macros (PSA_ALG_IS_xxx). For hash algorithms, validate the exact hash size, and validate xxx_GET_HASH macros on dependent algorithms. For MAC algorithms, validate the MAC size. For AEAD algorithms, validate the tag size. There is a separate test case for each HMAC algorithm, which is necessary because each has its own MAC size. For other hash-dependent algorithms, there is no interesting variation to test here, so only one hash gets tested.
2018-08-22 18:36:36 +02:00
Change the encoding of key types to have a parity bit Change the encoding of key types, EC curve families and DH group families to make the low-order bit a parity bit (with even parity). This ensures that distinct key type values always have a Hamming distance of at least 2, which makes it easier for implementations to resist single bit flips.
2019-12-04 18:58:44 +01:00
/* Check the parity of value.
Explain the purpose of check_parity Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2020-08-26 10:24:26 +02:00
*
* There are several numerical encodings for which the PSA Cryptography API
* specification deliberately defines encodings that all have the same
* parity. This way, a data glitch that flips one bit in the data cannot
* possibly turn a valid encoding into another valid encoding. Here in
* the tests, we check that the values (including Mbed TLS vendor-specific
* values) have the expected parity.
*
* The expected parity is even so that 0 is considered a valid encoding.
*
Fix the documentation of has_even_parity The documentation had the boolean meaning of the return value inverted. Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2020-08-26 12:14:37 +02:00
* Return a nonzero value if value has even parity and 0 otherwise. */
Clearer function name for parity check Return a name that more clearly returns nonzero=true=good, 0=bad. We'd normally expect check_xxx to return 0=pass, nonzero=fail so check_parity was a bad name. Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2020-08-26 11:16:50 +02:00
int has_even_parity( uint32_t value )
Change the encoding of key types to have a parity bit Change the encoding of key types, EC curve families and DH group families to make the low-order bit a parity bit (with even parity). This ensures that distinct key type values always have a Hamming distance of at least 2, which makes it easier for implementations to resist single bit flips.
2019-12-04 18:58:44 +01:00
{
value ^= value >> 16;
value ^= value >> 8;
value ^= value >> 4;
return( 0x9669 & 1 << ( value & 0xf ) );
}
#define TEST_PARITY( value ) \
Clearer function name for parity check Return a name that more clearly returns nonzero=true=good, 0=bad. We'd normally expect check_xxx to return 0=pass, nonzero=fail so check_parity was a bad name. Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2020-08-26 11:16:50 +02:00
TEST_ASSERT( has_even_parity( value ) )
Change the encoding of key types to have a parity bit Change the encoding of key types, EC curve families and DH group families to make the low-order bit a parity bit (with even parity). This ensures that distinct key type values always have a Hamming distance of at least 2, which makes it easier for implementations to resist single bit flips.
2019-12-04 18:58:44 +01:00
Algorithm validation tests For all algorithms, validate feature test macros (PSA_ALG_IS_xxx). For hash algorithms, validate the exact hash size, and validate xxx_GET_HASH macros on dependent algorithms. For MAC algorithms, validate the MAC size. For AEAD algorithms, validate the tag size. There is a separate test case for each HMAC algorithm, which is necessary because each has its own MAC size. For other hash-dependent algorithms, there is no interesting variation to test here, so only one hash gets tested.
2018-08-22 18:36:36 +02:00
void algorithm_classification( psa_algorithm_t alg, unsigned flags )
{
Ensure that all flags are actually tested At least twice, we added a classification flag but forgot to test it in the relevant test functions. Add some protection so that this doesn't happen again. In each classification category, put a macro xxx_FLAG_MASK_PLUS_ONE at the end. In the corresponding test function, keep track of the flags that are tested, and check that their mask is xxx_FLAG_MASK_PLUS_ONE - 1 which is all the bits of the previous flags set. Now, if we add a flag without testing it, the test TEST_EQUAL( classification_flags_tested, xxx_FLAG_MASK_PLUS_ONE - 1 ) will fail. It will also fail if we make the set of flag numbers non-consecutive, which is ok. This reveals that three algorithm flags had been added but not tested (in two separate occasions). Also, one key type flag that is no longer used by the library was still defined but not tested, which is not a test gap but is inconsistent. It's for DSA, which is relevant to the PSA encoding even if Mbed TLS doesn't implement it, so keep the flag and do test it. Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-11-03 14:18:08 +01:00
unsigned classification_flags_tested = 0;
TEST_CLASSIFICATION_MACRO( 1, ALG_IS_VENDOR_DEFINED, alg, flags );
TEST_CLASSIFICATION_MACRO( 1, ALG_IS_HMAC, alg, flags );
TEST_CLASSIFICATION_MACRO( 1, ALG_IS_BLOCK_CIPHER_MAC, alg, flags );
TEST_CLASSIFICATION_MACRO( 1, ALG_IS_STREAM_CIPHER, alg, flags );
TEST_CLASSIFICATION_MACRO( 1, ALG_IS_RSA_PKCS1V15_SIGN, alg, flags );
TEST_CLASSIFICATION_MACRO( 1, ALG_IS_RSA_PSS, alg, flags );
Fix test bug: some classification flags were not tested Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-11-03 13:56:47 +01:00
TEST_CLASSIFICATION_MACRO( 1, ALG_IS_RSA_PSS_ANY_SALT, alg, flags );
TEST_CLASSIFICATION_MACRO( 1, ALG_IS_RSA_PSS_STANDARD_SALT, alg, flags );
Ensure that all flags are actually tested At least twice, we added a classification flag but forgot to test it in the relevant test functions. Add some protection so that this doesn't happen again. In each classification category, put a macro xxx_FLAG_MASK_PLUS_ONE at the end. In the corresponding test function, keep track of the flags that are tested, and check that their mask is xxx_FLAG_MASK_PLUS_ONE - 1 which is all the bits of the previous flags set. Now, if we add a flag without testing it, the test TEST_EQUAL( classification_flags_tested, xxx_FLAG_MASK_PLUS_ONE - 1 ) will fail. It will also fail if we make the set of flag numbers non-consecutive, which is ok. This reveals that three algorithm flags had been added but not tested (in two separate occasions). Also, one key type flag that is no longer used by the library was still defined but not tested, which is not a test gap but is inconsistent. It's for DSA, which is relevant to the PSA encoding even if Mbed TLS doesn't implement it, so keep the flag and do test it. Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-11-03 14:18:08 +01:00
TEST_CLASSIFICATION_MACRO( 1, ALG_IS_DSA, alg, flags );
TEST_CLASSIFICATION_MACRO( PSA_ALG_IS_DSA( alg ),
ALG_DSA_IS_DETERMINISTIC, alg, flags );
TEST_CLASSIFICATION_MACRO( 1, ALG_IS_DETERMINISTIC_DSA, alg, flags );
TEST_CLASSIFICATION_MACRO( 1, ALG_IS_RANDOMIZED_DSA, alg, flags );
TEST_CLASSIFICATION_MACRO( 1, ALG_IS_ECDSA, alg, flags );
TEST_CLASSIFICATION_MACRO( PSA_ALG_IS_ECDSA( alg ),
ALG_ECDSA_IS_DETERMINISTIC, alg, flags );
TEST_CLASSIFICATION_MACRO( 1, ALG_IS_DETERMINISTIC_ECDSA, alg, flags );
TEST_CLASSIFICATION_MACRO( 1, ALG_IS_RANDOMIZED_ECDSA, alg, flags );
TEST_CLASSIFICATION_MACRO( 1, ALG_IS_HASH_EDDSA, alg, flags );
TEST_CLASSIFICATION_MACRO( 1, ALG_IS_SIGN_HASH, alg, flags );
TEST_CLASSIFICATION_MACRO( 1, ALG_IS_HASH_AND_SIGN, alg, flags );
TEST_CLASSIFICATION_MACRO( 1, ALG_IS_RSA_OAEP, alg, flags );
TEST_CLASSIFICATION_MACRO( 1, ALG_IS_HKDF, alg, flags );
TEST_CLASSIFICATION_MACRO( 1, ALG_IS_WILDCARD, alg, flags );
TEST_CLASSIFICATION_MACRO( 1, ALG_IS_ECDH, alg, flags );
TEST_CLASSIFICATION_MACRO( 1, ALG_IS_FFDH, alg, flags );
TEST_CLASSIFICATION_MACRO( 1, ALG_IS_RAW_KEY_AGREEMENT, alg, flags );
TEST_CLASSIFICATION_MACRO( 1, ALG_IS_AEAD_ON_BLOCK_CIPHER, alg, flags );
Fix test bug: some classification flags were not tested Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-11-03 13:56:47 +01:00
TEST_CLASSIFICATION_MACRO( 1, ALG_IS_TLS12_PRF, alg, flags );
TEST_CLASSIFICATION_MACRO( 1, ALG_IS_TLS12_PSK_TO_MS, alg, flags );
Ensure that all flags are actually tested At least twice, we added a classification flag but forgot to test it in the relevant test functions. Add some protection so that this doesn't happen again. In each classification category, put a macro xxx_FLAG_MASK_PLUS_ONE at the end. In the corresponding test function, keep track of the flags that are tested, and check that their mask is xxx_FLAG_MASK_PLUS_ONE - 1 which is all the bits of the previous flags set. Now, if we add a flag without testing it, the test TEST_EQUAL( classification_flags_tested, xxx_FLAG_MASK_PLUS_ONE - 1 ) will fail. It will also fail if we make the set of flag numbers non-consecutive, which is ok. This reveals that three algorithm flags had been added but not tested (in two separate occasions). Also, one key type flag that is no longer used by the library was still defined but not tested, which is not a test gap but is inconsistent. It's for DSA, which is relevant to the PSA encoding even if Mbed TLS doesn't implement it, so keep the flag and do test it. Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-11-03 14:18:08 +01:00
TEST_EQUAL( classification_flags_tested, ALG_FLAG_MASK_PLUS_ONE - 1 );
Algorithm validation tests For all algorithms, validate feature test macros (PSA_ALG_IS_xxx). For hash algorithms, validate the exact hash size, and validate xxx_GET_HASH macros on dependent algorithms. For MAC algorithms, validate the MAC size. For AEAD algorithms, validate the tag size. There is a separate test case for each HMAC algorithm, which is necessary because each has its own MAC size. For other hash-dependent algorithms, there is no interesting variation to test here, so only one hash gets tested.
2018-08-22 18:36:36 +02:00
exit: ;
}
Key type validation tests For all key types, validate feature test macros (PSA_KEY_TYPE_IS_xxx). For asymmetric keys (public key or key pair), validate the corresponding public/pair type. For ECC keys, validate GET_CURVE.
2018-08-22 18:43:09 +02:00
void key_type_classification( psa_key_type_t type, unsigned flags )
{
Ensure that all flags are actually tested At least twice, we added a classification flag but forgot to test it in the relevant test functions. Add some protection so that this doesn't happen again. In each classification category, put a macro xxx_FLAG_MASK_PLUS_ONE at the end. In the corresponding test function, keep track of the flags that are tested, and check that their mask is xxx_FLAG_MASK_PLUS_ONE - 1 which is all the bits of the previous flags set. Now, if we add a flag without testing it, the test TEST_EQUAL( classification_flags_tested, xxx_FLAG_MASK_PLUS_ONE - 1 ) will fail. It will also fail if we make the set of flag numbers non-consecutive, which is ok. This reveals that three algorithm flags had been added but not tested (in two separate occasions). Also, one key type flag that is no longer used by the library was still defined but not tested, which is not a test gap but is inconsistent. It's for DSA, which is relevant to the PSA encoding even if Mbed TLS doesn't implement it, so keep the flag and do test it. Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-11-03 14:18:08 +01:00
unsigned classification_flags_tested = 0;
Key type validation tests For all key types, validate feature test macros (PSA_KEY_TYPE_IS_xxx). For asymmetric keys (public key or key pair), validate the corresponding public/pair type. For ECC keys, validate GET_CURVE.
2018-08-22 18:43:09 +02:00
/* Macros tested based on the test case parameter */
Ensure that all flags are actually tested At least twice, we added a classification flag but forgot to test it in the relevant test functions. Add some protection so that this doesn't happen again. In each classification category, put a macro xxx_FLAG_MASK_PLUS_ONE at the end. In the corresponding test function, keep track of the flags that are tested, and check that their mask is xxx_FLAG_MASK_PLUS_ONE - 1 which is all the bits of the previous flags set. Now, if we add a flag without testing it, the test TEST_EQUAL( classification_flags_tested, xxx_FLAG_MASK_PLUS_ONE - 1 ) will fail. It will also fail if we make the set of flag numbers non-consecutive, which is ok. This reveals that three algorithm flags had been added but not tested (in two separate occasions). Also, one key type flag that is no longer used by the library was still defined but not tested, which is not a test gap but is inconsistent. It's for DSA, which is relevant to the PSA encoding even if Mbed TLS doesn't implement it, so keep the flag and do test it. Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-11-03 14:18:08 +01:00
TEST_CLASSIFICATION_MACRO( 1, KEY_TYPE_IS_VENDOR_DEFINED, type, flags );
TEST_CLASSIFICATION_MACRO( 1, KEY_TYPE_IS_UNSTRUCTURED, type, flags );
TEST_CLASSIFICATION_MACRO( 1, KEY_TYPE_IS_PUBLIC_KEY, type, flags );
TEST_CLASSIFICATION_MACRO( 1, KEY_TYPE_IS_KEY_PAIR, type, flags );
TEST_CLASSIFICATION_MACRO( 1, KEY_TYPE_IS_RSA, type, flags );
Fix test bug: some classification flags were not tested Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-11-03 13:56:47 +01:00
TEST_CLASSIFICATION_MACRO( 1, KEY_TYPE_IS_DSA, type, flags );
Ensure that all flags are actually tested At least twice, we added a classification flag but forgot to test it in the relevant test functions. Add some protection so that this doesn't happen again. In each classification category, put a macro xxx_FLAG_MASK_PLUS_ONE at the end. In the corresponding test function, keep track of the flags that are tested, and check that their mask is xxx_FLAG_MASK_PLUS_ONE - 1 which is all the bits of the previous flags set. Now, if we add a flag without testing it, the test TEST_EQUAL( classification_flags_tested, xxx_FLAG_MASK_PLUS_ONE - 1 ) will fail. It will also fail if we make the set of flag numbers non-consecutive, which is ok. This reveals that three algorithm flags had been added but not tested (in two separate occasions). Also, one key type flag that is no longer used by the library was still defined but not tested, which is not a test gap but is inconsistent. It's for DSA, which is relevant to the PSA encoding even if Mbed TLS doesn't implement it, so keep the flag and do test it. Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-11-03 14:18:08 +01:00
TEST_CLASSIFICATION_MACRO( 1, KEY_TYPE_IS_ECC, type, flags );
TEST_CLASSIFICATION_MACRO( 1, KEY_TYPE_IS_DH, type, flags );
TEST_EQUAL( classification_flags_tested, KEY_TYPE_FLAG_MASK_PLUS_ONE - 1 );
Key type validation tests For all key types, validate feature test macros (PSA_KEY_TYPE_IS_xxx). For asymmetric keys (public key or key pair), validate the corresponding public/pair type. For ECC keys, validate GET_CURVE.
2018-08-22 18:43:09 +02:00
/* Macros with derived semantics */
Use TEST_EQUAL(a,b) in preference to TEST_ASSERT(a==b) This commit is the result of the following command, followed by reindenting (but not wrapping lines): perl -00 -i -pe 's/^( *)TEST_ASSERT\(([^;=]*)(?: |\n *)==([^;=]*)\);$/${1}TEST_EQUAL($2,$3);/gm' tests/suites/test_suite_psa_*.function
2018-12-18 00:24:04 +01:00
TEST_EQUAL( PSA_KEY_TYPE_IS_ASYMMETRIC( type ),
( PSA_KEY_TYPE_IS_PUBLIC_KEY( type ) ||
Rename *KEYPAIR* to *KEY_PAIR* Be consistent with PUBLIC_KEY. perl -i -pe 's/KEYPAIR/KEY_PAIR/g' $(git ls-files)
2019-05-16 19:39:54 +02:00
PSA_KEY_TYPE_IS_KEY_PAIR( type ) ) );
TEST_EQUAL( PSA_KEY_TYPE_IS_ECC_KEY_PAIR( type ),
Use TEST_EQUAL(a,b) in preference to TEST_ASSERT(a==b) This commit is the result of the following command, followed by reindenting (but not wrapping lines): perl -00 -i -pe 's/^( *)TEST_ASSERT\(([^;=]*)(?: |\n *)==([^;=]*)\);$/${1}TEST_EQUAL($2,$3);/gm' tests/suites/test_suite_psa_*.function
2018-12-18 00:24:04 +01:00
( PSA_KEY_TYPE_IS_ECC( type ) &&
Rename *KEYPAIR* to *KEY_PAIR* Be consistent with PUBLIC_KEY. perl -i -pe 's/KEYPAIR/KEY_PAIR/g' $(git ls-files)
2019-05-16 19:39:54 +02:00
PSA_KEY_TYPE_IS_KEY_PAIR( type ) ) );
Use TEST_EQUAL(a,b) in preference to TEST_ASSERT(a==b) This commit is the result of the following command, followed by reindenting (but not wrapping lines): perl -00 -i -pe 's/^( *)TEST_ASSERT\(([^;=]*)(?: |\n *)==([^;=]*)\);$/${1}TEST_EQUAL($2,$3);/gm' tests/suites/test_suite_psa_*.function
2018-12-18 00:24:04 +01:00
TEST_EQUAL( PSA_KEY_TYPE_IS_ECC_PUBLIC_KEY( type ),
( PSA_KEY_TYPE_IS_ECC( type ) &&
PSA_KEY_TYPE_IS_PUBLIC_KEY( type ) ) );
Rename *KEYPAIR* to *KEY_PAIR* Be consistent with PUBLIC_KEY. perl -i -pe 's/KEYPAIR/KEY_PAIR/g' $(git ls-files)
2019-05-16 19:39:54 +02:00
TEST_EQUAL( PSA_KEY_TYPE_IS_DH_KEY_PAIR( type ),
Parametrize Diffie-Hellman keys by a group identifier Parametrize finite-field Diffie-Hellman key types with a DH group identifier, in the same way elliptic curve keys are parametrized with an EC curve identifier. Define the DH groups from the TLS registry (these are the groups from RFC 7919). Replicate the macro definitions and the metadata tests from elliptic curve identifiers to DH group identifiers. Define PSA_DH_GROUP_CUSTOM as an implementation-specific extension for which domain parameters are used to specify the group.
2019-05-16 12:55:35 +02:00
( PSA_KEY_TYPE_IS_DH( type ) &&
Rename *KEYPAIR* to *KEY_PAIR* Be consistent with PUBLIC_KEY. perl -i -pe 's/KEYPAIR/KEY_PAIR/g' $(git ls-files)
2019-05-16 19:39:54 +02:00
PSA_KEY_TYPE_IS_KEY_PAIR( type ) ) );
Parametrize Diffie-Hellman keys by a group identifier Parametrize finite-field Diffie-Hellman key types with a DH group identifier, in the same way elliptic curve keys are parametrized with an EC curve identifier. Define the DH groups from the TLS registry (these are the groups from RFC 7919). Replicate the macro definitions and the metadata tests from elliptic curve identifiers to DH group identifiers. Define PSA_DH_GROUP_CUSTOM as an implementation-specific extension for which domain parameters are used to specify the group.
2019-05-16 12:55:35 +02:00
TEST_EQUAL( PSA_KEY_TYPE_IS_DH_PUBLIC_KEY( type ),
( PSA_KEY_TYPE_IS_DH( type ) &&
PSA_KEY_TYPE_IS_PUBLIC_KEY( type ) ) );
Key type validation tests For all key types, validate feature test macros (PSA_KEY_TYPE_IS_xxx). For asymmetric keys (public key or key pair), validate the corresponding public/pair type. For ECC keys, validate GET_CURVE.
2018-08-22 18:43:09 +02:00
Change the encoding of key types to have a parity bit Change the encoding of key types, EC curve families and DH group families to make the low-order bit a parity bit (with even parity). This ensures that distinct key type values always have a Hamming distance of at least 2, which makes it easier for implementations to resist single bit flips.
2019-12-04 18:58:44 +01:00
TEST_PARITY( type );
Key type validation tests For all key types, validate feature test macros (PSA_KEY_TYPE_IS_xxx). For asymmetric keys (public key or key pair), validate the corresponding public/pair type. For ECC keys, validate GET_CURVE.
2018-08-22 18:43:09 +02:00
exit: ;
}
Add metadata tests for truncated MAC and short-tag AEAD
2018-09-24 22:09:46 +02:00
void mac_algorithm_core( psa_algorithm_t alg, int classification_flags,
psa_key_type_t key_type, size_t key_bits,
size_t length )
{
/* Algorithm classification */
TEST_ASSERT( ! PSA_ALG_IS_HASH( alg ) );
TEST_ASSERT( PSA_ALG_IS_MAC( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_CIPHER( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_AEAD( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_SIGN( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_ASYMMETRIC_ENCRYPTION( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_KEY_AGREEMENT( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_KEY_DERIVATION( alg ) );
PSA PAKE: Add J-PAKE to the interface Signed-off-by: Janos Follath <janos.follath@arm.com>
2021-03-21 16:11:01 +01:00
TEST_ASSERT( ! PSA_ALG_IS_PAKE( alg ) );
Add metadata tests for truncated MAC and short-tag AEAD
2018-09-24 22:09:46 +02:00
algorithm_classification( alg, classification_flags );
/* Length */
Rename output buffer size macros Rename existing support macros for output buffer sizes for PSA Crypto API 1.0.0 Signed-off-by: gabor-mezei-arm <gabor.mezei@arm.com>
2020-12-18 14:23:51 +01:00
TEST_EQUAL( length, PSA_MAC_LENGTH( key_type, key_bits, alg ) );
Add metadata tests for truncated MAC and short-tag AEAD
2018-09-24 22:09:46 +02:00
Make psa_get_mac_output_length testable and test it Signed-off-by: Steven Cooreman <steven.cooreman@silabs.com>
2021-03-02 21:31:17 +01:00
#if defined(MBEDTLS_TEST_HOOKS) && defined(MBEDTLS_PSA_CRYPTO_C)
Rework MAC algorithm / key type validation Reworked the validation of MAC algorithm with the used key type by introducing psa_mac_key_can_do, which guarantees that PSA_MAC_LENGTH can be called successfully after validation of the algorithm and key type. This means psa_get_mac_output_length is no longer required. Signed-off-by: Steven Cooreman <steven.cooreman@silabs.com>
2021-03-03 19:04:05 +01:00
PSA_ASSERT( psa_mac_key_can_do( alg, key_type ) );
Make psa_get_mac_output_length testable and test it Signed-off-by: Steven Cooreman <steven.cooreman@silabs.com>
2021-03-02 21:31:17 +01:00
#endif
Add metadata tests for truncated MAC and short-tag AEAD
2018-09-24 22:09:46 +02:00
exit: ;
}
void aead_algorithm_core( psa_algorithm_t alg, int classification_flags,
Update all uses of old AEAD output size macros Signed-off-by: Bence Szépkúti <bence.szepkuti@arm.com>
2021-03-19 18:46:15 +01:00
psa_key_type_t key_type, size_t key_bits,
Add metadata tests for truncated MAC and short-tag AEAD
2018-09-24 22:09:46 +02:00
size_t tag_length )
{
/* Algorithm classification */
TEST_ASSERT( ! PSA_ALG_IS_HASH( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_MAC( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_CIPHER( alg ) );
TEST_ASSERT( PSA_ALG_IS_AEAD( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_SIGN( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_ASYMMETRIC_ENCRYPTION( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_KEY_AGREEMENT( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_KEY_DERIVATION( alg ) );
PSA PAKE: Add J-PAKE to the interface Signed-off-by: Janos Follath <janos.follath@arm.com>
2021-03-21 16:11:01 +01:00
TEST_ASSERT( ! PSA_ALG_IS_PAKE( alg ) );
Add metadata tests for truncated MAC and short-tag AEAD
2018-09-24 22:09:46 +02:00
algorithm_classification( alg, classification_flags );
/* Tag length */
Update all uses of old AEAD output size macros Signed-off-by: Bence Szépkúti <bence.szepkuti@arm.com>
2021-03-19 18:46:15 +01:00
TEST_EQUAL( tag_length, PSA_AEAD_TAG_LENGTH( key_type, key_bits, alg ) );
Add metadata tests for truncated MAC and short-tag AEAD
2018-09-24 22:09:46 +02:00
exit: ;
}
Create infrastructure for metadata validation unit tests
2018-08-21 18:20:20 +02:00
/* END_HEADER */
/* BEGIN_DEPENDENCIES
tests: psa: Test PSA client-only code Signed-off-by: Ronald Cron <ronald.cron@arm.com>
2021-01-28 17:54:24 +01:00
* depends_on:MBEDTLS_PSA_CRYPTO_CLIENT
Create infrastructure for metadata validation unit tests
2018-08-21 18:20:20 +02:00
* END_DEPENDENCIES
*/
Algorithm validation tests For all algorithms, validate feature test macros (PSA_ALG_IS_xxx). For hash algorithms, validate the exact hash size, and validate xxx_GET_HASH macros on dependent algorithms. For MAC algorithms, validate the MAC size. For AEAD algorithms, validate the tag size. There is a separate test case for each HMAC algorithm, which is necessary because each has its own MAC size. For other hash-dependent algorithms, there is no interesting variation to test here, so only one hash gets tested.
2018-08-22 18:36:36 +02:00
/* BEGIN_CASE */
void hash_algorithm( int alg_arg, int length_arg )
{
psa_algorithm_t alg = alg_arg;
size_t length = length_arg;
psa_algorithm_t hmac_alg = PSA_ALG_HMAC( alg );
psa_algorithm_t rsa_pkcs1v15_sign_alg = PSA_ALG_RSA_PKCS1V15_SIGN( alg );
psa_algorithm_t rsa_pss_alg = PSA_ALG_RSA_PSS( alg );
psa_algorithm_t dsa_alg = PSA_ALG_DSA( alg );
psa_algorithm_t deterministic_dsa_alg = PSA_ALG_DETERMINISTIC_DSA( alg );
psa_algorithm_t ecdsa_alg = PSA_ALG_ECDSA( alg );
psa_algorithm_t deterministic_ecdsa_alg = PSA_ALG_DETERMINISTIC_ECDSA( alg );
psa_algorithm_t rsa_oaep_alg = PSA_ALG_RSA_OAEP( alg );
psa_algorithm_t hkdf_alg = PSA_ALG_HKDF( alg );
/* Algorithm classification */
TEST_ASSERT( PSA_ALG_IS_HASH( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_MAC( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_CIPHER( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_AEAD( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_SIGN( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_ASYMMETRIC_ENCRYPTION( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_KEY_AGREEMENT( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_KEY_DERIVATION( alg ) );
PSA PAKE: Add J-PAKE to the interface Signed-off-by: Janos Follath <janos.follath@arm.com>
2021-03-21 16:11:01 +01:00
TEST_ASSERT( ! PSA_ALG_IS_PAKE( alg ) );
Algorithm validation tests For all algorithms, validate feature test macros (PSA_ALG_IS_xxx). For hash algorithms, validate the exact hash size, and validate xxx_GET_HASH macros on dependent algorithms. For MAC algorithms, validate the MAC size. For AEAD algorithms, validate the tag size. There is a separate test case for each HMAC algorithm, which is necessary because each has its own MAC size. For other hash-dependent algorithms, there is no interesting variation to test here, so only one hash gets tested.
2018-08-22 18:36:36 +02:00
algorithm_classification( alg, 0 );
/* Dependent algorithms */
Use TEST_EQUAL(a,b) in preference to TEST_ASSERT(a==b) This commit is the result of the following command, followed by reindenting (but not wrapping lines): perl -00 -i -pe 's/^( *)TEST_ASSERT\(([^;=]*)(?: |\n *)==([^;=]*)\);$/${1}TEST_EQUAL($2,$3);/gm' tests/suites/test_suite_psa_*.function
2018-12-18 00:24:04 +01:00
TEST_EQUAL( PSA_ALG_HMAC_GET_HASH( hmac_alg ), alg );
TEST_EQUAL( PSA_ALG_SIGN_GET_HASH( rsa_pkcs1v15_sign_alg ), alg );
TEST_EQUAL( PSA_ALG_SIGN_GET_HASH( rsa_pss_alg ), alg );
TEST_EQUAL( PSA_ALG_SIGN_GET_HASH( dsa_alg ), alg );
TEST_EQUAL( PSA_ALG_SIGN_GET_HASH( deterministic_dsa_alg ), alg );
TEST_EQUAL( PSA_ALG_SIGN_GET_HASH( ecdsa_alg ), alg );
TEST_EQUAL( PSA_ALG_SIGN_GET_HASH( deterministic_ecdsa_alg ), alg );
TEST_EQUAL( PSA_ALG_RSA_OAEP_GET_HASH( rsa_oaep_alg ), alg );
TEST_EQUAL( PSA_ALG_HKDF_GET_HASH( hkdf_alg ), alg );
Algorithm validation tests For all algorithms, validate feature test macros (PSA_ALG_IS_xxx). For hash algorithms, validate the exact hash size, and validate xxx_GET_HASH macros on dependent algorithms. For MAC algorithms, validate the MAC size. For AEAD algorithms, validate the tag size. There is a separate test case for each HMAC algorithm, which is necessary because each has its own MAC size. For other hash-dependent algorithms, there is no interesting variation to test here, so only one hash gets tested.
2018-08-22 18:36:36 +02:00
Add block sizes to HMAC metadata validation tests
2018-09-17 14:13:26 +02:00
/* Hash length */
Rename output buffer size macros Rename existing support macros for output buffer sizes for PSA Crypto API 1.0.0 Signed-off-by: gabor-mezei-arm <gabor.mezei@arm.com>
2020-12-18 14:23:51 +01:00
TEST_EQUAL( length, PSA_HASH_LENGTH( alg ) );
Algorithm validation tests For all algorithms, validate feature test macros (PSA_ALG_IS_xxx). For hash algorithms, validate the exact hash size, and validate xxx_GET_HASH macros on dependent algorithms. For MAC algorithms, validate the MAC size. For AEAD algorithms, validate the tag size. There is a separate test case for each HMAC algorithm, which is necessary because each has its own MAC size. For other hash-dependent algorithms, there is no interesting variation to test here, so only one hash gets tested.
2018-08-22 18:36:36 +02:00
TEST_ASSERT( length <= PSA_HASH_MAX_SIZE );
}
/* END_CASE */
/* BEGIN_CASE */
void mac_algorithm( int alg_arg, int classification_flags,
int length_arg,
int key_type_arg, int key_bits_arg )
{
psa_algorithm_t alg = alg_arg;
size_t length = length_arg;
Add metadata tests for truncated MAC and short-tag AEAD
2018-09-24 22:09:46 +02:00
size_t n;
Algorithm validation tests For all algorithms, validate feature test macros (PSA_ALG_IS_xxx). For hash algorithms, validate the exact hash size, and validate xxx_GET_HASH macros on dependent algorithms. For MAC algorithms, validate the MAC size. For AEAD algorithms, validate the tag size. There is a separate test case for each HMAC algorithm, which is necessary because each has its own MAC size. For other hash-dependent algorithms, there is no interesting variation to test here, so only one hash gets tested.
2018-08-22 18:36:36 +02:00
size_t key_type = key_type_arg;
size_t key_bits = key_bits_arg;
Add metadata tests for truncated MAC and short-tag AEAD
2018-09-24 22:09:46 +02:00
mac_algorithm_core( alg, classification_flags,
key_type, key_bits, length );
Use TEST_EQUAL(a,b) in preference to TEST_ASSERT(a==b) This commit is the result of the following command, followed by reindenting (but not wrapping lines): perl -00 -i -pe 's/^( *)TEST_ASSERT\(([^;=]*)(?: |\n *)==([^;=]*)\);$/${1}TEST_EQUAL($2,$3);/gm' tests/suites/test_suite_psa_*.function
2018-12-18 00:24:04 +01:00
TEST_EQUAL( PSA_ALG_FULL_LENGTH_MAC( alg ), alg );
Algorithm validation tests For all algorithms, validate feature test macros (PSA_ALG_IS_xxx). For hash algorithms, validate the exact hash size, and validate xxx_GET_HASH macros on dependent algorithms. For MAC algorithms, validate the MAC size. For AEAD algorithms, validate the tag size. There is a separate test case for each HMAC algorithm, which is necessary because each has its own MAC size. For other hash-dependent algorithms, there is no interesting variation to test here, so only one hash gets tested.
2018-08-22 18:36:36 +02:00
TEST_ASSERT( length <= PSA_MAC_MAX_SIZE );
Add metadata tests for truncated MAC and short-tag AEAD
2018-09-24 22:09:46 +02:00
/* Truncated versions */
for( n = 1; n <= length; n++ )
{
psa_algorithm_t truncated_alg = PSA_ALG_TRUNCATED_MAC( alg, n );
mac_algorithm_core( truncated_alg, classification_flags,
key_type, key_bits, n );
Use TEST_EQUAL(a,b) in preference to TEST_ASSERT(a==b) This commit is the result of the following command, followed by reindenting (but not wrapping lines): perl -00 -i -pe 's/^( *)TEST_ASSERT\(([^;=]*)(?: |\n *)==([^;=]*)\);$/${1}TEST_EQUAL($2,$3);/gm' tests/suites/test_suite_psa_*.function
2018-12-18 00:24:04 +01:00
TEST_EQUAL( PSA_ALG_FULL_LENGTH_MAC( truncated_alg ), alg );
Add metadata tests for truncated MAC and short-tag AEAD
2018-09-24 22:09:46 +02:00
/* Check that calling PSA_ALG_TRUNCATED_MAC twice gives the length
* of the outer truncation (even if the outer length is smaller than
* the inner length). */
Use TEST_EQUAL(a,b) in preference to TEST_ASSERT(a==b) This commit is the result of the following command, followed by reindenting (but not wrapping lines): perl -00 -i -pe 's/^( *)TEST_ASSERT\(([^;=]*)(?: |\n *)==([^;=]*)\);$/${1}TEST_EQUAL($2,$3);/gm' tests/suites/test_suite_psa_*.function
2018-12-18 00:24:04 +01:00
TEST_EQUAL( PSA_ALG_TRUNCATED_MAC( truncated_alg, 1 ),
PSA_ALG_TRUNCATED_MAC( alg, 1 ) );
TEST_EQUAL( PSA_ALG_TRUNCATED_MAC( truncated_alg, length - 1 ),
PSA_ALG_TRUNCATED_MAC( alg, length - 1) );
TEST_EQUAL( PSA_ALG_TRUNCATED_MAC( truncated_alg, length ),
PSA_ALG_TRUNCATED_MAC( alg, length ) );
Add metadata tests intertwining truncated and at-least-length algos Signed-off-by: Steven Cooreman <steven.cooreman@silabs.com>
2021-02-25 11:22:03 +01:00
/* Check that calling PSA_ALG_TRUNCATED_MAC on an algorithm
* earlier constructed with PSA_ALG_AT_LEAST_THIS_LENGTH_MAC gives the
* length of the outer truncation (even if the outer length is smaller
* than the inner length). */
TEST_EQUAL( PSA_ALG_TRUNCATED_MAC(
PSA_ALG_AT_LEAST_THIS_LENGTH_MAC( truncated_alg, n ), 1 ),
PSA_ALG_TRUNCATED_MAC( alg, 1 ) );
TEST_EQUAL( PSA_ALG_TRUNCATED_MAC(
PSA_ALG_AT_LEAST_THIS_LENGTH_MAC( truncated_alg, n ), length - 1 ),
PSA_ALG_TRUNCATED_MAC( alg, length - 1) );
TEST_EQUAL( PSA_ALG_TRUNCATED_MAC(
PSA_ALG_AT_LEAST_THIS_LENGTH_MAC( truncated_alg, n ), length ),
PSA_ALG_TRUNCATED_MAC( alg, length ) );
Add metadata tests for truncated MAC and short-tag AEAD
2018-09-24 22:09:46 +02:00
}
Test _AT_LEAST_THIS_LENGTH macros in the PSA Crypto metadata test suite Signed-off-by: Steven Cooreman <steven.cooreman@silabs.com>
2021-02-23 10:21:58 +01:00
/* At-leat-this-length versions */
for( n = 1; n <= length; n++ )
{
psa_algorithm_t policy_alg = PSA_ALG_AT_LEAST_THIS_LENGTH_MAC( alg, n );
mac_algorithm_core( policy_alg, classification_flags | ALG_IS_WILDCARD,
key_type, key_bits, n );
TEST_EQUAL( PSA_ALG_FULL_LENGTH_MAC( policy_alg ), alg );
Fix typos & copy-paste errors Signed-off-by: Steven Cooreman <steven.cooreman@silabs.com>
2021-02-25 10:20:29 +01:00
/* Check that calling PSA_ALG_AT_LEAST_THIS_LENGTH_MAC twice gives the
* length of the outer truncation (even if the outer length is smaller
* than the inner length). */
Test _AT_LEAST_THIS_LENGTH macros in the PSA Crypto metadata test suite Signed-off-by: Steven Cooreman <steven.cooreman@silabs.com>
2021-02-23 10:21:58 +01:00
TEST_EQUAL( PSA_ALG_AT_LEAST_THIS_LENGTH_MAC( policy_alg, 1 ),
PSA_ALG_AT_LEAST_THIS_LENGTH_MAC( alg, 1 ) );
TEST_EQUAL( PSA_ALG_AT_LEAST_THIS_LENGTH_MAC( policy_alg, length - 1 ),
PSA_ALG_AT_LEAST_THIS_LENGTH_MAC( alg, length - 1) );
TEST_EQUAL( PSA_ALG_AT_LEAST_THIS_LENGTH_MAC( policy_alg, length ),
PSA_ALG_AT_LEAST_THIS_LENGTH_MAC( alg, length ) );
Add metadata tests intertwining truncated and at-least-length algos Signed-off-by: Steven Cooreman <steven.cooreman@silabs.com>
2021-02-25 11:22:03 +01:00
/* Check that calling PSA_ALG_AT_LEAST_THIS_LENGTH_MAC on an algorithm
* earlier constructed with PSA_ALG_TRUNCATED_MAC gives the length of
* the outer truncation (even if the outer length is smaller than the
* inner length). */
TEST_EQUAL( PSA_ALG_AT_LEAST_THIS_LENGTH_MAC(
PSA_ALG_TRUNCATED_MAC( policy_alg, n ), 1),
PSA_ALG_AT_LEAST_THIS_LENGTH_MAC( alg, 1 ) );
TEST_EQUAL( PSA_ALG_AT_LEAST_THIS_LENGTH_MAC(
PSA_ALG_TRUNCATED_MAC( policy_alg, n ), length - 1 ),
PSA_ALG_AT_LEAST_THIS_LENGTH_MAC( alg, length - 1) );
TEST_EQUAL( PSA_ALG_AT_LEAST_THIS_LENGTH_MAC(
PSA_ALG_TRUNCATED_MAC( policy_alg, n ), length ),
PSA_ALG_AT_LEAST_THIS_LENGTH_MAC( alg, length ) );
Test _AT_LEAST_THIS_LENGTH macros in the PSA Crypto metadata test suite Signed-off-by: Steven Cooreman <steven.cooreman@silabs.com>
2021-02-23 10:21:58 +01:00
}
Algorithm validation tests For all algorithms, validate feature test macros (PSA_ALG_IS_xxx). For hash algorithms, validate the exact hash size, and validate xxx_GET_HASH macros on dependent algorithms. For MAC algorithms, validate the MAC size. For AEAD algorithms, validate the tag size. There is a separate test case for each HMAC algorithm, which is necessary because each has its own MAC size. For other hash-dependent algorithms, there is no interesting variation to test here, so only one hash gets tested.
2018-08-22 18:36:36 +02:00
}
/* END_CASE */
Add block sizes to HMAC metadata validation tests
2018-09-17 14:13:26 +02:00
/* BEGIN_CASE */
void hmac_algorithm( int alg_arg,
int length_arg,
int block_size_arg )
{
psa_algorithm_t alg = alg_arg;
psa_algorithm_t hash_alg = PSA_ALG_HMAC_GET_HASH( alg );
size_t block_size = block_size_arg;
Add metadata tests for truncated MAC and short-tag AEAD
2018-09-24 22:09:46 +02:00
size_t length = length_arg;
size_t n;
Add block sizes to HMAC metadata validation tests
2018-09-17 14:13:26 +02:00
TEST_ASSERT( PSA_ALG_IS_HASH( hash_alg ) );
Use TEST_EQUAL(a,b) in preference to TEST_ASSERT(a==b) This commit is the result of the following command, followed by reindenting (but not wrapping lines): perl -00 -i -pe 's/^( *)TEST_ASSERT\(([^;=]*)(?: |\n *)==([^;=]*)\);$/${1}TEST_EQUAL($2,$3);/gm' tests/suites/test_suite_psa_*.function
2018-12-18 00:24:04 +01:00
TEST_EQUAL( PSA_ALG_HMAC( hash_alg ), alg );
Add block sizes to HMAC metadata validation tests
2018-09-17 14:13:26 +02:00
Test PSA_HASH_BLOCK_LENGTH Only tested for algorithms for which we support HMAC, since that's all we use PSA_HASH_BLOCK_LENGTH for at the moment. Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-09-22 14:47:26 +02:00
TEST_ASSERT( block_size == PSA_HASH_BLOCK_LENGTH( alg ) );
Add block sizes to HMAC metadata validation tests
2018-09-17 14:13:26 +02:00
TEST_ASSERT( block_size <= PSA_HMAC_MAX_HASH_BLOCK_SIZE );
Add metadata tests for truncated MAC and short-tag AEAD
2018-09-24 22:09:46 +02:00
test_mac_algorithm( alg_arg, ALG_IS_HMAC, length,
PSA_KEY_TYPE_HMAC, PSA_BYTES_TO_BITS( length ) );
for( n = 1; n <= length; n++ )
{
psa_algorithm_t truncated_alg = PSA_ALG_TRUNCATED_MAC( alg, n );
Use TEST_EQUAL(a,b) in preference to TEST_ASSERT(a==b) This commit is the result of the following command, followed by reindenting (but not wrapping lines): perl -00 -i -pe 's/^( *)TEST_ASSERT\(([^;=]*)(?: |\n *)==([^;=]*)\);$/${1}TEST_EQUAL($2,$3);/gm' tests/suites/test_suite_psa_*.function
2018-12-18 00:24:04 +01:00
TEST_EQUAL( PSA_ALG_HMAC_GET_HASH( truncated_alg ), hash_alg );
Add metadata tests for truncated MAC and short-tag AEAD
2018-09-24 22:09:46 +02:00
}
Add block sizes to HMAC metadata validation tests
2018-09-17 14:13:26 +02:00
}
/* END_CASE */
Algorithm validation tests For all algorithms, validate feature test macros (PSA_ALG_IS_xxx). For hash algorithms, validate the exact hash size, and validate xxx_GET_HASH macros on dependent algorithms. For MAC algorithms, validate the MAC size. For AEAD algorithms, validate the tag size. There is a separate test case for each HMAC algorithm, which is necessary because each has its own MAC size. For other hash-dependent algorithms, there is no interesting variation to test here, so only one hash gets tested.
2018-08-22 18:36:36 +02:00
/* BEGIN_CASE */
void cipher_algorithm( int alg_arg, int classification_flags )
{
psa_algorithm_t alg = alg_arg;
/* Algorithm classification */
TEST_ASSERT( ! PSA_ALG_IS_HASH( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_MAC( alg ) );
TEST_ASSERT( PSA_ALG_IS_CIPHER( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_AEAD( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_SIGN( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_ASYMMETRIC_ENCRYPTION( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_KEY_AGREEMENT( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_KEY_DERIVATION( alg ) );
PSA PAKE: Add J-PAKE to the interface Signed-off-by: Janos Follath <janos.follath@arm.com>
2021-03-21 16:11:01 +01:00
TEST_ASSERT( ! PSA_ALG_IS_PAKE( alg ) );
Algorithm validation tests For all algorithms, validate feature test macros (PSA_ALG_IS_xxx). For hash algorithms, validate the exact hash size, and validate xxx_GET_HASH macros on dependent algorithms. For MAC algorithms, validate the MAC size. For AEAD algorithms, validate the tag size. There is a separate test case for each HMAC algorithm, which is necessary because each has its own MAC size. For other hash-dependent algorithms, there is no interesting variation to test here, so only one hash gets tested.
2018-08-22 18:36:36 +02:00
algorithm_classification( alg, classification_flags );
}
/* END_CASE */
/* BEGIN_CASE */
void aead_algorithm( int alg_arg, int classification_flags,
Update all uses of old AEAD output size macros Signed-off-by: Bence Szépkúti <bence.szepkuti@arm.com>
2021-03-19 18:46:15 +01:00
int tag_length_arg,
int key_type_arg, int key_bits_arg )
Algorithm validation tests For all algorithms, validate feature test macros (PSA_ALG_IS_xxx). For hash algorithms, validate the exact hash size, and validate xxx_GET_HASH macros on dependent algorithms. For MAC algorithms, validate the MAC size. For AEAD algorithms, validate the tag size. There is a separate test case for each HMAC algorithm, which is necessary because each has its own MAC size. For other hash-dependent algorithms, there is no interesting variation to test here, so only one hash gets tested.
2018-08-22 18:36:36 +02:00
{
psa_algorithm_t alg = alg_arg;
size_t tag_length = tag_length_arg;
Add metadata tests for truncated MAC and short-tag AEAD
2018-09-24 22:09:46 +02:00
size_t n;
Update all uses of old AEAD output size macros Signed-off-by: Bence Szépkúti <bence.szepkuti@arm.com>
2021-03-19 18:46:15 +01:00
psa_key_type_t key_type = key_type_arg;
size_t key_bits = key_bits_arg;
Algorithm validation tests For all algorithms, validate feature test macros (PSA_ALG_IS_xxx). For hash algorithms, validate the exact hash size, and validate xxx_GET_HASH macros on dependent algorithms. For MAC algorithms, validate the MAC size. For AEAD algorithms, validate the tag size. There is a separate test case for each HMAC algorithm, which is necessary because each has its own MAC size. For other hash-dependent algorithms, there is no interesting variation to test here, so only one hash gets tested.
2018-08-22 18:36:36 +02:00
Update all uses of old AEAD output size macros Signed-off-by: Bence Szépkúti <bence.szepkuti@arm.com>
2021-03-19 18:46:15 +01:00
aead_algorithm_core( alg, classification_flags,
key_type, key_bits, tag_length );
Algorithm validation tests For all algorithms, validate feature test macros (PSA_ALG_IS_xxx). For hash algorithms, validate the exact hash size, and validate xxx_GET_HASH macros on dependent algorithms. For MAC algorithms, validate the MAC size. For AEAD algorithms, validate the tag size. There is a separate test case for each HMAC algorithm, which is necessary because each has its own MAC size. For other hash-dependent algorithms, there is no interesting variation to test here, so only one hash gets tested.
2018-08-22 18:36:36 +02:00
Add metadata tests for truncated MAC and short-tag AEAD
2018-09-24 22:09:46 +02:00
/* Truncated versions */
for( n = 1; n <= tag_length; n++ )
{
Rename AEAD tag length macros This brings them in line with PSA Crypto API 1.0.0 PSA_ALG_AEAD_WITH_DEFAULT_TAG_LENGTH -> PSA_ALG_AEAD_WITH_DEFAULT_LENGTH_TAG PSA_ALG_AEAD_WITH_TAG_LENGTH -> PSA_ALG_AEAD_WITH_SHORTENED_TAG Signed-off-by: Bence Szépkúti <bence.szepkuti@arm.com>
2020-12-16 11:36:46 +01:00
psa_algorithm_t truncated_alg = PSA_ALG_AEAD_WITH_SHORTENED_TAG( alg, n );
Update all uses of old AEAD output size macros Signed-off-by: Bence Szépkúti <bence.szepkuti@arm.com>
2021-03-19 18:46:15 +01:00
aead_algorithm_core( truncated_alg, classification_flags,
key_type, key_bits, n );
Rename AEAD tag length macros This brings them in line with PSA Crypto API 1.0.0 PSA_ALG_AEAD_WITH_DEFAULT_TAG_LENGTH -> PSA_ALG_AEAD_WITH_DEFAULT_LENGTH_TAG PSA_ALG_AEAD_WITH_TAG_LENGTH -> PSA_ALG_AEAD_WITH_SHORTENED_TAG Signed-off-by: Bence Szépkúti <bence.szepkuti@arm.com>
2020-12-16 11:36:46 +01:00
TEST_EQUAL( PSA_ALG_AEAD_WITH_DEFAULT_LENGTH_TAG( truncated_alg ),
Rewrap some lines after the macro changes Change the way some lines are wrapped to cut at a more logical place. This commit mainly rewrites multi-line calls to TEST_EQUAL, and also a few calls to PSA_ASSERT.
2018-12-18 00:33:25 +01:00
alg );
Fix typos & copy-paste errors Signed-off-by: Steven Cooreman <steven.cooreman@silabs.com>
2021-02-25 10:20:29 +01:00
/* Check that calling PSA_ALG_AEAD_WITH_SHORTENED_TAG twice gives
Add metadata tests for truncated MAC and short-tag AEAD
2018-09-24 22:09:46 +02:00
* the length of the outer truncation (even if the outer length is
* smaller than the inner length). */
Rename AEAD tag length macros This brings them in line with PSA Crypto API 1.0.0 PSA_ALG_AEAD_WITH_DEFAULT_TAG_LENGTH -> PSA_ALG_AEAD_WITH_DEFAULT_LENGTH_TAG PSA_ALG_AEAD_WITH_TAG_LENGTH -> PSA_ALG_AEAD_WITH_SHORTENED_TAG Signed-off-by: Bence Szépkúti <bence.szepkuti@arm.com>
2020-12-16 11:36:46 +01:00
TEST_EQUAL( PSA_ALG_AEAD_WITH_SHORTENED_TAG( truncated_alg, 1 ),
PSA_ALG_AEAD_WITH_SHORTENED_TAG( alg, 1 ) );
TEST_EQUAL( PSA_ALG_AEAD_WITH_SHORTENED_TAG( truncated_alg, tag_length - 1 ),
PSA_ALG_AEAD_WITH_SHORTENED_TAG( alg, tag_length - 1) );
TEST_EQUAL( PSA_ALG_AEAD_WITH_SHORTENED_TAG( truncated_alg, tag_length ),
PSA_ALG_AEAD_WITH_SHORTENED_TAG( alg, tag_length ) );
Add metadata tests intertwining truncated and at-least-length algos Signed-off-by: Steven Cooreman <steven.cooreman@silabs.com>
2021-02-25 11:22:03 +01:00
/* Check that calling PSA_ALG_AEAD_WITH_SHORTENED_TAG on an algorithm
* earlier constructed with PSA_ALG_AEAD_WITH_AT_LEAST_THIS_LENGTH_TAG
* gives the length of the outer truncation (even if the outer length is
* smaller than the inner length). */
TEST_EQUAL( PSA_ALG_AEAD_WITH_SHORTENED_TAG(
PSA_ALG_AEAD_WITH_AT_LEAST_THIS_LENGTH_TAG( truncated_alg, n ), 1 ),
PSA_ALG_AEAD_WITH_SHORTENED_TAG( alg, 1 ) );
TEST_EQUAL( PSA_ALG_AEAD_WITH_SHORTENED_TAG(
PSA_ALG_AEAD_WITH_AT_LEAST_THIS_LENGTH_TAG( truncated_alg, n ), tag_length - 1 ),
PSA_ALG_AEAD_WITH_SHORTENED_TAG( alg, tag_length - 1) );
TEST_EQUAL( PSA_ALG_AEAD_WITH_SHORTENED_TAG(
PSA_ALG_AEAD_WITH_AT_LEAST_THIS_LENGTH_TAG( truncated_alg, n ), tag_length ),
PSA_ALG_AEAD_WITH_SHORTENED_TAG( alg, tag_length ) );
Add metadata tests for truncated MAC and short-tag AEAD
2018-09-24 22:09:46 +02:00
}
Test _AT_LEAST_THIS_LENGTH macros in the PSA Crypto metadata test suite Signed-off-by: Steven Cooreman <steven.cooreman@silabs.com>
2021-02-23 10:21:58 +01:00
/* At-leat-this-length versions */
for( n = 1; n <= tag_length; n++ )
{
psa_algorithm_t policy_alg = PSA_ALG_AEAD_WITH_AT_LEAST_THIS_LENGTH_TAG( alg, n );
Update all uses of old AEAD output size macros Signed-off-by: Bence Szépkúti <bence.szepkuti@arm.com>
2021-03-19 18:46:15 +01:00
aead_algorithm_core( policy_alg, classification_flags | ALG_IS_WILDCARD,
key_type, key_bits, n );
Test _AT_LEAST_THIS_LENGTH macros in the PSA Crypto metadata test suite Signed-off-by: Steven Cooreman <steven.cooreman@silabs.com>
2021-02-23 10:21:58 +01:00
TEST_EQUAL( PSA_ALG_AEAD_WITH_DEFAULT_LENGTH_TAG( policy_alg ),
alg );
Fix typos & copy-paste errors Signed-off-by: Steven Cooreman <steven.cooreman@silabs.com>
2021-02-25 10:20:29 +01:00
/* Check that calling PSA_ALG_AEAD_WITH_AT_LEAST_THIS_LENGTH_TAG twice
* gives the length of the outer truncation (even if the outer length is
Test _AT_LEAST_THIS_LENGTH macros in the PSA Crypto metadata test suite Signed-off-by: Steven Cooreman <steven.cooreman@silabs.com>
2021-02-23 10:21:58 +01:00
* smaller than the inner length). */
TEST_EQUAL( PSA_ALG_AEAD_WITH_AT_LEAST_THIS_LENGTH_TAG( policy_alg, 1 ),
PSA_ALG_AEAD_WITH_AT_LEAST_THIS_LENGTH_TAG( alg, 1 ) );
TEST_EQUAL( PSA_ALG_AEAD_WITH_AT_LEAST_THIS_LENGTH_TAG( policy_alg, tag_length - 1 ),
PSA_ALG_AEAD_WITH_AT_LEAST_THIS_LENGTH_TAG( alg, tag_length - 1) );
TEST_EQUAL( PSA_ALG_AEAD_WITH_AT_LEAST_THIS_LENGTH_TAG( policy_alg, tag_length ),
PSA_ALG_AEAD_WITH_AT_LEAST_THIS_LENGTH_TAG( alg, tag_length ) );
Add metadata tests intertwining truncated and at-least-length algos Signed-off-by: Steven Cooreman <steven.cooreman@silabs.com>
2021-02-25 11:22:03 +01:00
/* Check that calling PSA_ALG_AEAD_WITH_AT_LEAST_THIS_LENGTH_TAG on an
* algorithm earlier constructed with PSA_ALG_AEAD_WITH_SHORTENED_TAG
* gives the length of the outer truncation (even if the outer length is
* smaller than the inner length). */
TEST_EQUAL( PSA_ALG_AEAD_WITH_AT_LEAST_THIS_LENGTH_TAG(
PSA_ALG_AEAD_WITH_SHORTENED_TAG( policy_alg, n ), 1),
PSA_ALG_AEAD_WITH_AT_LEAST_THIS_LENGTH_TAG( alg, 1 ) );
TEST_EQUAL( PSA_ALG_AEAD_WITH_AT_LEAST_THIS_LENGTH_TAG(
PSA_ALG_AEAD_WITH_SHORTENED_TAG( policy_alg, n ), tag_length - 1 ),
PSA_ALG_AEAD_WITH_AT_LEAST_THIS_LENGTH_TAG( alg, tag_length - 1) );
TEST_EQUAL( PSA_ALG_AEAD_WITH_AT_LEAST_THIS_LENGTH_TAG(
PSA_ALG_AEAD_WITH_SHORTENED_TAG( policy_alg, n ), tag_length ),
PSA_ALG_AEAD_WITH_AT_LEAST_THIS_LENGTH_TAG( alg, tag_length ) );
Test _AT_LEAST_THIS_LENGTH macros in the PSA Crypto metadata test suite Signed-off-by: Steven Cooreman <steven.cooreman@silabs.com>
2021-02-23 10:21:58 +01:00
}
Algorithm validation tests For all algorithms, validate feature test macros (PSA_ALG_IS_xxx). For hash algorithms, validate the exact hash size, and validate xxx_GET_HASH macros on dependent algorithms. For MAC algorithms, validate the MAC size. For AEAD algorithms, validate the tag size. There is a separate test case for each HMAC algorithm, which is necessary because each has its own MAC size. For other hash-dependent algorithms, there is no interesting variation to test here, so only one hash gets tested.
2018-08-22 18:36:36 +02:00
}
/* END_CASE */
/* BEGIN_CASE */
void asymmetric_signature_algorithm( int alg_arg, int classification_flags )
{
psa_algorithm_t alg = alg_arg;
/* Algorithm classification */
TEST_ASSERT( ! PSA_ALG_IS_HASH( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_MAC( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_CIPHER( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_AEAD( alg ) );
TEST_ASSERT( PSA_ALG_IS_SIGN( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_ASYMMETRIC_ENCRYPTION( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_KEY_AGREEMENT( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_KEY_DERIVATION( alg ) );
PSA PAKE: Add J-PAKE to the interface Signed-off-by: Janos Follath <janos.follath@arm.com>
2021-03-21 16:11:01 +01:00
TEST_ASSERT( ! PSA_ALG_IS_PAKE( alg ) );
Algorithm validation tests For all algorithms, validate feature test macros (PSA_ALG_IS_xxx). For hash algorithms, validate the exact hash size, and validate xxx_GET_HASH macros on dependent algorithms. For MAC algorithms, validate the MAC size. For AEAD algorithms, validate the tag size. There is a separate test case for each HMAC algorithm, which is necessary because each has its own MAC size. For other hash-dependent algorithms, there is no interesting variation to test here, so only one hash gets tested.
2018-08-22 18:36:36 +02:00
algorithm_classification( alg, classification_flags );
}
/* END_CASE */
Add a hash wildcard value for hash-and-sign algorithm You can use PSA_ALG_ANY_HASH to build the algorithm value for a hash-and-sign algorithm in a policy. Then the policy allows usage with this hash-and-sign family with any hash. Test that PSA_ALG_ANY_HASH-based policies allow a specific hash, but not a different hash-and-sign family. Test that PSA_ALG_ANY_HASH is not valid for operations, only in policies.
2019-01-14 16:06:39 +01:00
/* BEGIN_CASE */
void asymmetric_signature_wildcard( int alg_arg, int classification_flags )
{
Add PSA_ALG_IS_HASH_AND_SIGN to the metadata tests The status of signature wildcards with respect to PSA_ALG_IS_HASH_AND_SIGN is unclear in the specification. A wildcard is usually instantiated with a specific hash, making the implementation hash-and-sign, but it could also be instantiated with a non-hash-and-sign algorithm. For the time being, go with what's currently implemented, which is that they are considered hash-and-sign. Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-09-22 18:12:31 +02:00
classification_flags |= ALG_IS_WILDCARD;
classification_flags |= ALG_IS_SIGN_HASH;
classification_flags |= ALG_IS_HASH_AND_SIGN;
Add a hash wildcard value for hash-and-sign algorithm You can use PSA_ALG_ANY_HASH to build the algorithm value for a hash-and-sign algorithm in a policy. Then the policy allows usage with this hash-and-sign family with any hash. Test that PSA_ALG_ANY_HASH-based policies allow a specific hash, but not a different hash-and-sign family. Test that PSA_ALG_ANY_HASH is not valid for operations, only in policies.
2019-01-14 16:06:39 +01:00
test_asymmetric_signature_algorithm( alg_arg, classification_flags );
/* Any failure of this test function comes from
* asymmetric_signature_algorithm. Pacify -Werror=unused-label. */
goto exit;
}
/* END_CASE */
Algorithm validation tests For all algorithms, validate feature test macros (PSA_ALG_IS_xxx). For hash algorithms, validate the exact hash size, and validate xxx_GET_HASH macros on dependent algorithms. For MAC algorithms, validate the MAC size. For AEAD algorithms, validate the tag size. There is a separate test case for each HMAC algorithm, which is necessary because each has its own MAC size. For other hash-dependent algorithms, there is no interesting variation to test here, so only one hash gets tested.
2018-08-22 18:36:36 +02:00
/* BEGIN_CASE */
void asymmetric_encryption_algorithm( int alg_arg, int classification_flags )
{
psa_algorithm_t alg = alg_arg;
/* Algorithm classification */
TEST_ASSERT( ! PSA_ALG_IS_HASH( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_MAC( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_CIPHER( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_AEAD( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_SIGN( alg ) );
TEST_ASSERT( PSA_ALG_IS_ASYMMETRIC_ENCRYPTION( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_KEY_AGREEMENT( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_KEY_DERIVATION( alg ) );
PSA PAKE: Add J-PAKE to the interface Signed-off-by: Janos Follath <janos.follath@arm.com>
2021-03-21 16:11:01 +01:00
TEST_ASSERT( ! PSA_ALG_IS_PAKE( alg ) );
Algorithm validation tests For all algorithms, validate feature test macros (PSA_ALG_IS_xxx). For hash algorithms, validate the exact hash size, and validate xxx_GET_HASH macros on dependent algorithms. For MAC algorithms, validate the MAC size. For AEAD algorithms, validate the tag size. There is a separate test case for each HMAC algorithm, which is necessary because each has its own MAC size. For other hash-dependent algorithms, there is no interesting variation to test here, so only one hash gets tested.
2018-08-22 18:36:36 +02:00
algorithm_classification( alg, classification_flags );
}
/* END_CASE */
/* BEGIN_CASE */
New algorithm category: key selection A key selection algorithm is similar to a key derivation algorithm in that it takes a secret input and produces a secret output stream. However, unlike key derivation algorithms, there is no expectation that the input cannot be reconstructed from the output. Key selection algorithms are exclusively meant to be used on the output of a key agreement algorithm to select chunks of the shared secret.
2018-09-18 11:52:10 +02:00
void key_derivation_algorithm( int alg_arg, int classification_flags )
Algorithm validation tests For all algorithms, validate feature test macros (PSA_ALG_IS_xxx). For hash algorithms, validate the exact hash size, and validate xxx_GET_HASH macros on dependent algorithms. For MAC algorithms, validate the MAC size. For AEAD algorithms, validate the tag size. There is a separate test case for each HMAC algorithm, which is necessary because each has its own MAC size. For other hash-dependent algorithms, there is no interesting variation to test here, so only one hash gets tested.
2018-08-22 18:36:36 +02:00
{
psa_algorithm_t alg = alg_arg;
Simplify the encoding of key agreement algorithms Get rid of "key selection" algorithms (of which there was only one: raw key selection). Encode key agreement by combining a raw key agreement with a KDF, rather than passing the KDF as an argument of a key agreement macro.
2019-01-18 16:44:49 +01:00
psa_algorithm_t ecdh_alg = PSA_ALG_KEY_AGREEMENT( PSA_ALG_ECDH, alg );
psa_algorithm_t ffdh_alg = PSA_ALG_KEY_AGREEMENT( PSA_ALG_FFDH, alg );
Algorithm validation tests For all algorithms, validate feature test macros (PSA_ALG_IS_xxx). For hash algorithms, validate the exact hash size, and validate xxx_GET_HASH macros on dependent algorithms. For MAC algorithms, validate the MAC size. For AEAD algorithms, validate the tag size. There is a separate test case for each HMAC algorithm, which is necessary because each has its own MAC size. For other hash-dependent algorithms, there is no interesting variation to test here, so only one hash gets tested.
2018-08-22 18:36:36 +02:00
/* Algorithm classification */
TEST_ASSERT( ! PSA_ALG_IS_HASH( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_MAC( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_CIPHER( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_AEAD( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_SIGN( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_ASYMMETRIC_ENCRYPTION( alg ) );
New algorithm category: key selection A key selection algorithm is similar to a key derivation algorithm in that it takes a secret input and produces a secret output stream. However, unlike key derivation algorithms, there is no expectation that the input cannot be reconstructed from the output. Key selection algorithms are exclusively meant to be used on the output of a key agreement algorithm to select chunks of the shared secret.
2018-09-18 11:52:10 +02:00
TEST_ASSERT( ! PSA_ALG_IS_KEY_AGREEMENT( alg ) );
TEST_ASSERT( PSA_ALG_IS_KEY_DERIVATION( alg ) );
PSA PAKE: Add J-PAKE to the interface Signed-off-by: Janos Follath <janos.follath@arm.com>
2021-03-21 16:11:01 +01:00
TEST_ASSERT( ! PSA_ALG_IS_PAKE( alg ) );
Algorithm validation tests For all algorithms, validate feature test macros (PSA_ALG_IS_xxx). For hash algorithms, validate the exact hash size, and validate xxx_GET_HASH macros on dependent algorithms. For MAC algorithms, validate the MAC size. For AEAD algorithms, validate the tag size. There is a separate test case for each HMAC algorithm, which is necessary because each has its own MAC size. For other hash-dependent algorithms, there is no interesting variation to test here, so only one hash gets tested.
2018-08-22 18:36:36 +02:00
algorithm_classification( alg, classification_flags );
Key agreement: macros for finite-field Diffie-Hellman, ECDH Declare macros to represent key agreement algorithms.
2018-09-18 11:54:43 +02:00
/* Check combinations with key agreements */
Simplify the encoding of key agreement algorithms Get rid of "key selection" algorithms (of which there was only one: raw key selection). Encode key agreement by combining a raw key agreement with a KDF, rather than passing the KDF as an argument of a key agreement macro.
2019-01-18 16:44:49 +01:00
TEST_ASSERT( PSA_ALG_IS_KEY_AGREEMENT( ecdh_alg ) );
TEST_ASSERT( PSA_ALG_IS_KEY_AGREEMENT( ffdh_alg ) );
TEST_EQUAL( PSA_ALG_KEY_AGREEMENT_GET_KDF( ecdh_alg ), alg );
TEST_EQUAL( PSA_ALG_KEY_AGREEMENT_GET_KDF( ffdh_alg ), alg );
New algorithm category: key selection A key selection algorithm is similar to a key derivation algorithm in that it takes a secret input and produces a secret output stream. However, unlike key derivation algorithms, there is no expectation that the input cannot be reconstructed from the output. Key selection algorithms are exclusively meant to be used on the output of a key agreement algorithm to select chunks of the shared secret.
2018-09-18 11:52:10 +02:00
}
/* END_CASE */
/* BEGIN_CASE */
void key_agreement_algorithm( int alg_arg, int classification_flags,
Simplify the encoding of key agreement algorithms Get rid of "key selection" algorithms (of which there was only one: raw key selection). Encode key agreement by combining a raw key agreement with a KDF, rather than passing the KDF as an argument of a key agreement macro.
2019-01-18 16:44:49 +01:00
int ka_alg_arg, int kdf_alg_arg )
New algorithm category: key selection A key selection algorithm is similar to a key derivation algorithm in that it takes a secret input and produces a secret output stream. However, unlike key derivation algorithms, there is no expectation that the input cannot be reconstructed from the output. Key selection algorithms are exclusively meant to be used on the output of a key agreement algorithm to select chunks of the shared secret.
2018-09-18 11:52:10 +02:00
{
psa_algorithm_t alg = alg_arg;
Simplify the encoding of key agreement algorithms Get rid of "key selection" algorithms (of which there was only one: raw key selection). Encode key agreement by combining a raw key agreement with a KDF, rather than passing the KDF as an argument of a key agreement macro.
2019-01-18 16:44:49 +01:00
psa_algorithm_t actual_ka_alg = PSA_ALG_KEY_AGREEMENT_GET_BASE( alg );
psa_algorithm_t expected_ka_alg = ka_alg_arg;
psa_algorithm_t actual_kdf_alg = PSA_ALG_KEY_AGREEMENT_GET_KDF( alg );
psa_algorithm_t expected_kdf_alg = kdf_alg_arg;
New algorithm category: key selection A key selection algorithm is similar to a key derivation algorithm in that it takes a secret input and produces a secret output stream. However, unlike key derivation algorithms, there is no expectation that the input cannot be reconstructed from the output. Key selection algorithms are exclusively meant to be used on the output of a key agreement algorithm to select chunks of the shared secret.
2018-09-18 11:52:10 +02:00
/* Algorithm classification */
TEST_ASSERT( ! PSA_ALG_IS_HASH( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_MAC( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_CIPHER( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_AEAD( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_SIGN( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_ASYMMETRIC_ENCRYPTION( alg ) );
TEST_ASSERT( PSA_ALG_IS_KEY_AGREEMENT( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_KEY_DERIVATION( alg ) );
PSA PAKE: Add J-PAKE to the interface Signed-off-by: Janos Follath <janos.follath@arm.com>
2021-03-21 16:11:01 +01:00
TEST_ASSERT( ! PSA_ALG_IS_PAKE( alg ) );
Algorithm validation tests For all algorithms, validate feature test macros (PSA_ALG_IS_xxx). For hash algorithms, validate the exact hash size, and validate xxx_GET_HASH macros on dependent algorithms. For MAC algorithms, validate the MAC size. For AEAD algorithms, validate the tag size. There is a separate test case for each HMAC algorithm, which is necessary because each has its own MAC size. For other hash-dependent algorithms, there is no interesting variation to test here, so only one hash gets tested.
2018-08-22 18:36:36 +02:00
algorithm_classification( alg, classification_flags );
New algorithm category: key selection A key selection algorithm is similar to a key derivation algorithm in that it takes a secret input and produces a secret output stream. However, unlike key derivation algorithms, there is no expectation that the input cannot be reconstructed from the output. Key selection algorithms are exclusively meant to be used on the output of a key agreement algorithm to select chunks of the shared secret.
2018-09-18 11:52:10 +02:00
/* Shared secret derivation properties */
Simplify the encoding of key agreement algorithms Get rid of "key selection" algorithms (of which there was only one: raw key selection). Encode key agreement by combining a raw key agreement with a KDF, rather than passing the KDF as an argument of a key agreement macro.
2019-01-18 16:44:49 +01:00
TEST_EQUAL( actual_ka_alg, expected_ka_alg );
TEST_EQUAL( actual_kdf_alg, expected_kdf_alg );
Algorithm validation tests For all algorithms, validate feature test macros (PSA_ALG_IS_xxx). For hash algorithms, validate the exact hash size, and validate xxx_GET_HASH macros on dependent algorithms. For MAC algorithms, validate the MAC size. For AEAD algorithms, validate the tag size. There is a separate test case for each HMAC algorithm, which is necessary because each has its own MAC size. For other hash-dependent algorithms, there is no interesting variation to test here, so only one hash gets tested.
2018-08-22 18:36:36 +02:00
}
/* END_CASE */
PSA PAKE: Add J-PAKE to the interface Signed-off-by: Janos Follath <janos.follath@arm.com>
2021-03-21 16:11:01 +01:00
/* BEGIN_CASE */
void pake_algorithm( int alg_arg )
{
psa_algorithm_t alg = alg_arg;
/* Algorithm classification */
TEST_ASSERT( ! PSA_ALG_IS_HASH( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_MAC( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_CIPHER( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_AEAD( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_SIGN( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_ASYMMETRIC_ENCRYPTION( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_KEY_AGREEMENT( alg ) );
TEST_ASSERT( ! PSA_ALG_IS_KEY_DERIVATION( alg ) );
TEST_ASSERT( PSA_ALG_IS_PAKE( alg ) );
}
/* END_CASE */
Key type validation tests For all key types, validate feature test macros (PSA_KEY_TYPE_IS_xxx). For asymmetric keys (public key or key pair), validate the corresponding public/pair type. For ECC keys, validate GET_CURVE.
2018-08-22 18:43:09 +02:00
/* BEGIN_CASE */
void key_type( int type_arg, int classification_flags )
{
psa_key_type_t type = type_arg;
key_type_classification( type, classification_flags );
/* For asymmetric types, check the corresponding pair/public type */
if( classification_flags & KEY_TYPE_IS_PUBLIC_KEY )
{
Rename *KEYPAIR* to *KEY_PAIR* Be consistent with PUBLIC_KEY. perl -i -pe 's/KEYPAIR/KEY_PAIR/g' $(git ls-files)
2019-05-16 19:39:54 +02:00
psa_key_type_t pair_type = PSA_KEY_TYPE_KEY_PAIR_OF_PUBLIC_KEY( type );
TEST_EQUAL( PSA_KEY_TYPE_PUBLIC_KEY_OF_KEY_PAIR( pair_type ), type );
Key type validation tests For all key types, validate feature test macros (PSA_KEY_TYPE_IS_xxx). For asymmetric keys (public key or key pair), validate the corresponding public/pair type. For ECC keys, validate GET_CURVE.
2018-08-22 18:43:09 +02:00
key_type_classification( pair_type,
( classification_flags
& ~KEY_TYPE_IS_PUBLIC_KEY )
Rename *KEYPAIR* to *KEY_PAIR* Be consistent with PUBLIC_KEY. perl -i -pe 's/KEYPAIR/KEY_PAIR/g' $(git ls-files)
2019-05-16 19:39:54 +02:00
| KEY_TYPE_IS_KEY_PAIR );
TEST_EQUAL( PSA_KEY_TYPE_PUBLIC_KEY_OF_KEY_PAIR( type ), type );
Key type validation tests For all key types, validate feature test macros (PSA_KEY_TYPE_IS_xxx). For asymmetric keys (public key or key pair), validate the corresponding public/pair type. For ECC keys, validate GET_CURVE.
2018-08-22 18:43:09 +02:00
}
Rename *KEYPAIR* to *KEY_PAIR* Be consistent with PUBLIC_KEY. perl -i -pe 's/KEYPAIR/KEY_PAIR/g' $(git ls-files)
2019-05-16 19:39:54 +02:00
if( classification_flags & KEY_TYPE_IS_KEY_PAIR )
Key type validation tests For all key types, validate feature test macros (PSA_KEY_TYPE_IS_xxx). For asymmetric keys (public key or key pair), validate the corresponding public/pair type. For ECC keys, validate GET_CURVE.
2018-08-22 18:43:09 +02:00
{
Rename *KEYPAIR* to *KEY_PAIR* Be consistent with PUBLIC_KEY. perl -i -pe 's/KEYPAIR/KEY_PAIR/g' $(git ls-files)
2019-05-16 19:39:54 +02:00
psa_key_type_t public_type = PSA_KEY_TYPE_PUBLIC_KEY_OF_KEY_PAIR( type );
TEST_EQUAL( PSA_KEY_TYPE_KEY_PAIR_OF_PUBLIC_KEY( public_type ), type );
Key type validation tests For all key types, validate feature test macros (PSA_KEY_TYPE_IS_xxx). For asymmetric keys (public key or key pair), validate the corresponding public/pair type. For ECC keys, validate GET_CURVE.
2018-08-22 18:43:09 +02:00
key_type_classification( public_type,
( classification_flags
Rename *KEYPAIR* to *KEY_PAIR* Be consistent with PUBLIC_KEY. perl -i -pe 's/KEYPAIR/KEY_PAIR/g' $(git ls-files)
2019-05-16 19:39:54 +02:00
& ~KEY_TYPE_IS_KEY_PAIR )
Key type validation tests For all key types, validate feature test macros (PSA_KEY_TYPE_IS_xxx). For asymmetric keys (public key or key pair), validate the corresponding public/pair type. For ECC keys, validate GET_CURVE.
2018-08-22 18:43:09 +02:00
| KEY_TYPE_IS_PUBLIC_KEY );
Rename *KEYPAIR* to *KEY_PAIR* Be consistent with PUBLIC_KEY. perl -i -pe 's/KEYPAIR/KEY_PAIR/g' $(git ls-files)
2019-05-16 19:39:54 +02:00
TEST_EQUAL( PSA_KEY_TYPE_KEY_PAIR_OF_PUBLIC_KEY( type ), type );
Key type validation tests For all key types, validate feature test macros (PSA_KEY_TYPE_IS_xxx). For asymmetric keys (public key or key pair), validate the corresponding public/pair type. For ECC keys, validate GET_CURVE.
2018-08-22 18:43:09 +02:00
}
}
/* END_CASE */
Test the block size for symmetric keys Also insist on their category. Fix a missing implementation of PSA_BLOCK_CIPHER_BLOCK_SIZE for ChaCha20.
2019-12-02 17:26:44 +01:00
/* BEGIN_CASE */
Improve Code Style add space between bracket and a newline that had occured through changes but do not match the original file style Signed-off-by: Joe Subbiani <joe.subbiani@arm.com>
2021-07-09 13:03:06 +02:00
void block_cipher_key_type( int type_arg, int block_size_arg )
Test the block size for symmetric keys Also insist on their category. Fix a missing implementation of PSA_BLOCK_CIPHER_BLOCK_SIZE for ChaCha20.
2019-12-02 17:26:44 +01:00
{
psa_key_type_t type = type_arg;
size_t block_size = block_size_arg;
test_key_type( type_arg, KEY_TYPE_IS_UNSTRUCTURED );
TEST_EQUAL( type & PSA_KEY_TYPE_CATEGORY_MASK,
PSA_KEY_TYPE_CATEGORY_SYMMETRIC );
Remove test cases and update power 2 check The power 2 check now uses a looping bit shift to try match with the block sizes and will escape the loop when appropriate The test cases, as pointed out by Gilles, could be harmful in the future and testing a test case is not generally necessary Signed-off-by: Joe Subbiani <joe.subbiani@arm.com>
2021-07-08 16:32:52 +02:00
TEST_EQUAL( PSA_BLOCK_CIPHER_BLOCK_LENGTH( type ), block_size );
Extend psa_crypto_metadata tests to check for powers of 2 Add a check to ensure the block_size is or is not a power of 2 Add a new parameter to verify the expected pass/fail when a block_size is or is not a power of 2. Add new sets of input data to verify these tests Fixes #4228 Signed-off-by: Joe Subbiani <joe.subbiani@arm.com>
2021-07-06 11:42:54 +02:00
Remove trailing whitespace Signed-off-by: Joe Subbiani <joe.subbiani@arm.com>
2021-07-15 10:03:51 +02:00
/* Check that the block size is a power of 2. This is required, at least,
Simplify the test and description Previously the check was convoluted. This has been simplified and given a more appropriate suggestion as per gilles suggestion Signed-off-by: Joe Subbiani <joe.subbiani@arm.com>
2021-07-14 16:16:09 +02:00
for PSA_ROUND_UP_TO_MULTIPLE(block_size, length) in crypto_sizes.h. */
TEST_ASSERT( ( ( block_size - 1 ) & block_size ) == 0 );
Test the block size for symmetric keys Also insist on their category. Fix a missing implementation of PSA_BLOCK_CIPHER_BLOCK_SIZE for ChaCha20.
2019-12-02 17:26:44 +01:00
}
/* END_CASE */
/* BEGIN_CASE */
void stream_cipher_key_type( int type_arg )
{
psa_key_type_t type = type_arg;
test_key_type( type_arg, KEY_TYPE_IS_UNSTRUCTURED );
TEST_EQUAL( type & PSA_KEY_TYPE_CATEGORY_MASK,
PSA_KEY_TYPE_CATEGORY_SYMMETRIC );
Rename output buffer size macros Rename existing support macros for output buffer sizes for PSA Crypto API 1.0.0 Signed-off-by: gabor-mezei-arm <gabor.mezei@arm.com>
2020-12-18 14:23:51 +01:00
TEST_EQUAL( PSA_BLOCK_CIPHER_BLOCK_LENGTH( type ), 1 );
Test the block size for symmetric keys Also insist on their category. Fix a missing implementation of PSA_BLOCK_CIPHER_BLOCK_SIZE for ChaCha20.
2019-12-02 17:26:44 +01:00
}
/* END_CASE */
Update dependencies in PSA test code Replace MBEDTLS_* config options for which there is an associated PSA_WANT_* to the PSA_WANT_* one. That way the tests are also run when the dependency is provided by a driver. Signed-off-by: Ronald Cron <ronald.cron@arm.com>
2021-03-18 18:50:08 +01:00
/* BEGIN_CASE depends_on:PSA_KEY_TYPE_ECC_PUBLIC_KEY:PSA_KEY_TYPE_ECC_KEY_PAIR */
Define EC curve family constants Define constants for ECC curve families and DH group families. These constants have 0x0000 in the lower 16 bits of the key type. Support these constants in the implementation and in the PSA metadata tests. Switch the slot management and secure element driver HAL tests to the new curve encodings. This requires SE driver code to become slightly more clever when figuring out the bit-size of an imported EC key since it now needs to take the data size into account. Switch some documentation to the new encodings. Remove the macro PSA_ECC_CURVE_BITS which can no longer be implemented.
2019-12-03 17:24:19 +01:00
void ecc_key_family( int curve_arg )
Key type validation tests For all key types, validate feature test macros (PSA_KEY_TYPE_IS_xxx). For asymmetric keys (public key or key pair), validate the corresponding public/pair type. For ECC keys, validate GET_CURVE.
2018-08-22 18:43:09 +02:00
{
Rename ECC Family Macros According to PSA Spec Rename PSA_ECC_CURVE_xxx to PSA_ECC_FAMILY_xxx, also rename PSA_KEY_TYPE_GET_CURVE to PSA_KEY_TYPE_ECC_GET_FAMILY and rename psa_ecc_curve_t to psa_ecc_family_t. Old defines are provided in include/crypto_compat.h for backward compatibility. Signed-off-by: Paul Elliott <paul.elliott@arm.com>
2020-06-02 18:19:28 +02:00
psa_ecc_family_t curve = curve_arg;
Key type validation tests For all key types, validate feature test macros (PSA_KEY_TYPE_IS_xxx). For asymmetric keys (public key or key pair), validate the corresponding public/pair type. For ECC keys, validate GET_CURVE.
2018-08-22 18:43:09 +02:00
psa_key_type_t public_type = PSA_KEY_TYPE_ECC_PUBLIC_KEY( curve );
Rename *KEYPAIR* to *KEY_PAIR* Be consistent with PUBLIC_KEY. perl -i -pe 's/KEYPAIR/KEY_PAIR/g' $(git ls-files)
2019-05-16 19:39:54 +02:00
psa_key_type_t pair_type = PSA_KEY_TYPE_ECC_KEY_PAIR( curve );
Key type validation tests For all key types, validate feature test macros (PSA_KEY_TYPE_IS_xxx). For asymmetric keys (public key or key pair), validate the corresponding public/pair type. For ECC keys, validate GET_CURVE.
2018-08-22 18:43:09 +02:00
Fix parity tests to actually fail the test on error Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2020-08-25 22:30:31 +02:00
TEST_PARITY( curve );
Change the encoding of key types to have a parity bit Change the encoding of key types, EC curve families and DH group families to make the low-order bit a parity bit (with even parity). This ensures that distinct key type values always have a Hamming distance of at least 2, which makes it easier for implementations to resist single bit flips.
2019-12-04 18:58:44 +01:00
Key type validation tests For all key types, validate feature test macros (PSA_KEY_TYPE_IS_xxx). For asymmetric keys (public key or key pair), validate the corresponding public/pair type. For ECC keys, validate GET_CURVE.
2018-08-22 18:43:09 +02:00
test_key_type( public_type, KEY_TYPE_IS_ECC | KEY_TYPE_IS_PUBLIC_KEY );
Rename *KEYPAIR* to *KEY_PAIR* Be consistent with PUBLIC_KEY. perl -i -pe 's/KEYPAIR/KEY_PAIR/g' $(git ls-files)
2019-05-16 19:39:54 +02:00
test_key_type( pair_type, KEY_TYPE_IS_ECC | KEY_TYPE_IS_KEY_PAIR );
Key type validation tests For all key types, validate feature test macros (PSA_KEY_TYPE_IS_xxx). For asymmetric keys (public key or key pair), validate the corresponding public/pair type. For ECC keys, validate GET_CURVE.
2018-08-22 18:43:09 +02:00
Rename ECC Family Macros According to PSA Spec Rename PSA_ECC_CURVE_xxx to PSA_ECC_FAMILY_xxx, also rename PSA_KEY_TYPE_GET_CURVE to PSA_KEY_TYPE_ECC_GET_FAMILY and rename psa_ecc_curve_t to psa_ecc_family_t. Old defines are provided in include/crypto_compat.h for backward compatibility. Signed-off-by: Paul Elliott <paul.elliott@arm.com>
2020-06-02 18:19:28 +02:00
TEST_EQUAL( PSA_KEY_TYPE_ECC_GET_FAMILY( public_type ), curve );
TEST_EQUAL( PSA_KEY_TYPE_ECC_GET_FAMILY( pair_type ), curve );
Define EC curve family constants Define constants for ECC curve families and DH group families. These constants have 0x0000 in the lower 16 bits of the key type. Support these constants in the implementation and in the PSA metadata tests. Switch the slot management and secure element driver HAL tests to the new curve encodings. This requires SE driver code to become slightly more clever when figuring out the bit-size of an imported EC key since it now needs to take the data size into account. Switch some documentation to the new encodings. Remove the macro PSA_ECC_CURVE_BITS which can no longer be implemented.
2019-12-03 17:24:19 +01:00
}
/* END_CASE */
Parametrize Diffie-Hellman keys by a group identifier Parametrize finite-field Diffie-Hellman key types with a DH group identifier, in the same way elliptic curve keys are parametrized with an EC curve identifier. Define the DH groups from the TLS registry (these are the groups from RFC 7919). Replicate the macro definitions and the metadata tests from elliptic curve identifiers to DH group identifiers. Define PSA_DH_GROUP_CUSTOM as an implementation-specific extension for which domain parameters are used to specify the group.
2019-05-16 12:55:35 +02:00
/* BEGIN_CASE depends_on:MBEDTLS_DHM_C */
Define EC curve family constants Define constants for ECC curve families and DH group families. These constants have 0x0000 in the lower 16 bits of the key type. Support these constants in the implementation and in the PSA metadata tests. Switch the slot management and secure element driver HAL tests to the new curve encodings. This requires SE driver code to become slightly more clever when figuring out the bit-size of an imported EC key since it now needs to take the data size into account. Switch some documentation to the new encodings. Remove the macro PSA_ECC_CURVE_BITS which can no longer be implemented.
2019-12-03 17:24:19 +01:00
void dh_key_family( int group_arg )
Parametrize Diffie-Hellman keys by a group identifier Parametrize finite-field Diffie-Hellman key types with a DH group identifier, in the same way elliptic curve keys are parametrized with an EC curve identifier. Define the DH groups from the TLS registry (these are the groups from RFC 7919). Replicate the macro definitions and the metadata tests from elliptic curve identifiers to DH group identifiers. Define PSA_DH_GROUP_CUSTOM as an implementation-specific extension for which domain parameters are used to specify the group.
2019-05-16 12:55:35 +02:00
{
Rename DH Family Macros According to PSA Spec Rename PSA_DH_GROUP_xxx to PSA_DH_FAMILY_xxx, also rename PSA_KEY_TYPE_GET_GROUP to PSA_KEY_TYPE_DH_GET_FAMILY and rename psa_dh_group_t to psa_dh_family_t. Old defines are provided in include/crypto_compat.h for backward compatibility. Signed-off-by: Paul Elliott <paul.elliott@arm.com>
2020-06-03 16:17:39 +02:00
psa_dh_family_t group = group_arg;
Parametrize Diffie-Hellman keys by a group identifier Parametrize finite-field Diffie-Hellman key types with a DH group identifier, in the same way elliptic curve keys are parametrized with an EC curve identifier. Define the DH groups from the TLS registry (these are the groups from RFC 7919). Replicate the macro definitions and the metadata tests from elliptic curve identifiers to DH group identifiers. Define PSA_DH_GROUP_CUSTOM as an implementation-specific extension for which domain parameters are used to specify the group.
2019-05-16 12:55:35 +02:00
psa_key_type_t public_type = PSA_KEY_TYPE_DH_PUBLIC_KEY( group );
Rename *KEYPAIR* to *KEY_PAIR* Be consistent with PUBLIC_KEY. perl -i -pe 's/KEYPAIR/KEY_PAIR/g' $(git ls-files)
2019-05-16 19:39:54 +02:00
psa_key_type_t pair_type = PSA_KEY_TYPE_DH_KEY_PAIR( group );
Parametrize Diffie-Hellman keys by a group identifier Parametrize finite-field Diffie-Hellman key types with a DH group identifier, in the same way elliptic curve keys are parametrized with an EC curve identifier. Define the DH groups from the TLS registry (these are the groups from RFC 7919). Replicate the macro definitions and the metadata tests from elliptic curve identifiers to DH group identifiers. Define PSA_DH_GROUP_CUSTOM as an implementation-specific extension for which domain parameters are used to specify the group.
2019-05-16 12:55:35 +02:00
Fix parity tests to actually fail the test on error Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2020-08-25 22:30:31 +02:00
TEST_PARITY( group );
Change the encoding of key types to have a parity bit Change the encoding of key types, EC curve families and DH group families to make the low-order bit a parity bit (with even parity). This ensures that distinct key type values always have a Hamming distance of at least 2, which makes it easier for implementations to resist single bit flips.
2019-12-04 18:58:44 +01:00
Parametrize Diffie-Hellman keys by a group identifier Parametrize finite-field Diffie-Hellman key types with a DH group identifier, in the same way elliptic curve keys are parametrized with an EC curve identifier. Define the DH groups from the TLS registry (these are the groups from RFC 7919). Replicate the macro definitions and the metadata tests from elliptic curve identifiers to DH group identifiers. Define PSA_DH_GROUP_CUSTOM as an implementation-specific extension for which domain parameters are used to specify the group.
2019-05-16 12:55:35 +02:00
test_key_type( public_type, KEY_TYPE_IS_DH | KEY_TYPE_IS_PUBLIC_KEY );
Rename *KEYPAIR* to *KEY_PAIR* Be consistent with PUBLIC_KEY. perl -i -pe 's/KEYPAIR/KEY_PAIR/g' $(git ls-files)
2019-05-16 19:39:54 +02:00
test_key_type( pair_type, KEY_TYPE_IS_DH | KEY_TYPE_IS_KEY_PAIR );
Parametrize Diffie-Hellman keys by a group identifier Parametrize finite-field Diffie-Hellman key types with a DH group identifier, in the same way elliptic curve keys are parametrized with an EC curve identifier. Define the DH groups from the TLS registry (these are the groups from RFC 7919). Replicate the macro definitions and the metadata tests from elliptic curve identifiers to DH group identifiers. Define PSA_DH_GROUP_CUSTOM as an implementation-specific extension for which domain parameters are used to specify the group.
2019-05-16 12:55:35 +02:00
Rename DH Family Macros According to PSA Spec Rename PSA_DH_GROUP_xxx to PSA_DH_FAMILY_xxx, also rename PSA_KEY_TYPE_GET_GROUP to PSA_KEY_TYPE_DH_GET_FAMILY and rename psa_dh_group_t to psa_dh_family_t. Old defines are provided in include/crypto_compat.h for backward compatibility. Signed-off-by: Paul Elliott <paul.elliott@arm.com>
2020-06-03 16:17:39 +02:00
TEST_EQUAL( PSA_KEY_TYPE_DH_GET_FAMILY( public_type ), group );
TEST_EQUAL( PSA_KEY_TYPE_DH_GET_FAMILY( pair_type ), group );
Define EC curve family constants Define constants for ECC curve families and DH group families. These constants have 0x0000 in the lower 16 bits of the key type. Support these constants in the implementation and in the PSA metadata tests. Switch the slot management and secure element driver HAL tests to the new curve encodings. This requires SE driver code to become slightly more clever when figuring out the bit-size of an imported EC key since it now needs to take the data size into account. Switch some documentation to the new encodings. Remove the macro PSA_ECC_CURVE_BITS which can no longer be implemented.
2019-12-03 17:24:19 +01:00
}
/* END_CASE */
Add lifetime metadata tests Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-04-21 20:03:53 +02:00
/* BEGIN_CASE */
void lifetime( int lifetime_arg, int classification_flags,
int persistence_arg, int location_arg )
{
psa_key_lifetime_t lifetime = lifetime_arg;
psa_key_persistence_t persistence = persistence_arg;
psa_key_location_t location = location_arg;
unsigned flags = classification_flags;
Ensure that all flags are actually tested At least twice, we added a classification flag but forgot to test it in the relevant test functions. Add some protection so that this doesn't happen again. In each classification category, put a macro xxx_FLAG_MASK_PLUS_ONE at the end. In the corresponding test function, keep track of the flags that are tested, and check that their mask is xxx_FLAG_MASK_PLUS_ONE - 1 which is all the bits of the previous flags set. Now, if we add a flag without testing it, the test TEST_EQUAL( classification_flags_tested, xxx_FLAG_MASK_PLUS_ONE - 1 ) will fail. It will also fail if we make the set of flag numbers non-consecutive, which is ok. This reveals that three algorithm flags had been added but not tested (in two separate occasions). Also, one key type flag that is no longer used by the library was still defined but not tested, which is not a test gap but is inconsistent. It's for DSA, which is relevant to the PSA encoding even if Mbed TLS doesn't implement it, so keep the flag and do test it. Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-11-03 14:18:08 +01:00
unsigned classification_flags_tested = 0;
Add lifetime metadata tests Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-04-21 20:03:53 +02:00
Ensure that all flags are actually tested At least twice, we added a classification flag but forgot to test it in the relevant test functions. Add some protection so that this doesn't happen again. In each classification category, put a macro xxx_FLAG_MASK_PLUS_ONE at the end. In the corresponding test function, keep track of the flags that are tested, and check that their mask is xxx_FLAG_MASK_PLUS_ONE - 1 which is all the bits of the previous flags set. Now, if we add a flag without testing it, the test TEST_EQUAL( classification_flags_tested, xxx_FLAG_MASK_PLUS_ONE - 1 ) will fail. It will also fail if we make the set of flag numbers non-consecutive, which is ok. This reveals that three algorithm flags had been added but not tested (in two separate occasions). Also, one key type flag that is no longer used by the library was still defined but not tested, which is not a test gap but is inconsistent. It's for DSA, which is relevant to the PSA encoding even if Mbed TLS doesn't implement it, so keep the flag and do test it. Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-11-03 14:18:08 +01:00
TEST_CLASSIFICATION_MACRO( 1, KEY_LIFETIME_IS_VOLATILE, lifetime, flags );
TEST_CLASSIFICATION_MACRO( 1, KEY_LIFETIME_IS_READ_ONLY, lifetime, flags );
TEST_EQUAL( classification_flags_tested,
KEY_LIFETIME_FLAG_MASK_PLUS_ONE - 1 );
Add lifetime metadata tests Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
2021-04-21 20:03:53 +02:00
TEST_EQUAL( PSA_KEY_LIFETIME_GET_PERSISTENCE( lifetime ), persistence );
TEST_EQUAL( PSA_KEY_LIFETIME_GET_LOCATION( lifetime ), location );
}
/* END_CASE */
Reference in a new issue Copy permalink