Commit graph

263 commits

Author SHA1 Message Date
Gilles Peskine
86a440b638 Reject non-byte-aligned RSA keys
On key import and key generation, for RSA, reject key sizes that are
not a multiple of 8. Such keys are not well-supported in Mbed TLS and
are hardly ever used in practice.

The previous commit removed support for non-byte-aligned keys at the
PSA level. This commit actively rejects such keys and adds
corresponding tests (test keys generated with "openssl genrsa").
2018-11-12 19:14:18 +01:00
Gilles Peskine
aac64a2839 Remove support for non-byte-aligned RSA keys
Remove the need for an extra function mbedtls_rsa_get_bitlen. Use
mbedtls_rsa_get_len, which is only correct for keys whose size is a
multiple of 8. Key sizes that aren't a multiple of 8 are extremely
rarely used, so in practice this is not a problematic limitation.
2018-11-12 18:37:42 +01:00
Gilles Peskine
52b9018cf7 psa_export_key: for raw-byte keys, zero the end of the output buffer
Skip all writing to the target buffer if its size is 0, since in this
case the pointer might be invalid and this would cause the calls to
memcpy and memset to have undefined behavior.
2018-10-31 14:10:06 +01:00
Gilles Peskine
188c71e382 Private EC key format: change to raw secret value (export)
Change the import/export format of private elliptic curve keys from
RFC 5915 to the raw secret value. This commit updates the export code.
2018-10-31 14:10:06 +01:00
Gilles Peskine
f76aa7789b Private EC key format: change to raw secret value (doc, import)
Change the import/export format of private elliptic curve keys from
RFC 5915 to the raw secret value. This commit updates the format
specification and the import code, but not the export code.
2018-10-31 14:10:06 +01:00
Gilles Peskine
e0e9c7c417 New macro PSA_ALG_FULL_LENGTH_MAC
Provide a documented way of constructing the full-length MAC algorithm
from a truncated version.
2018-10-17 18:30:47 +02:00
Gilles Peskine
57fbdb1939 Use a public macro for AEAD tag length variations
Avoid depending on the encoding of algorithms inside psa_crypto.c.
2018-10-17 18:30:47 +02:00
Gilles Peskine
f8a8fe60f8 Fix memory leak with AEAD with non-default tag lengths
When freeing the key context, choose the context format based on the
base algorithm value stored in the operation object.
2018-10-17 13:54:48 +02:00
Gilles Peskine
8cac2e628e Translate GCM_BAD_INPUT to INVALID_ARGUMENT, not NOT_SUPPORTED 2018-10-17 13:54:48 +02:00
Gilles Peskine
99b7d6b700 Wipe sensitive data in psa_mac_verify_finish
Wipe the whole MAC intermediate buffer, not just the requested MAC
size. With truncated MAC algorithms, the requested MAC size may be
smaller than what is written to the intermediate buffer.
2018-10-17 13:54:47 +02:00
Gilles Peskine
87b0ac49f8 Fix possible buffer overread in psa_mac_finish_internal (CMAC) 2018-10-17 13:54:47 +02:00
Gilles Peskine
6d72ff9e79 Document that the minimum truncated MAC length is implementation-defined 2018-10-17 13:54:47 +02:00
Gilles Peskine
23cc2ff9a8 Add support for non-default-tag-size AEAD (CCM and GCM) 2018-10-08 14:42:11 +02:00
Gilles Peskine
edf9a6576d Refactor AEAD setup into a common function
There was a lot of repetition between psa_aead_encrypt and
psa_aead_decrypt. Refactor the code into a new function psa_aead_setup.
The new code should behave identically except that in some cases where
multiple error conditions apply, the code may now return a different
error code.

Internally, I rearranged some of the code:

* I removed a check that the key type was in CATEGORY_SYMMETRIC because
  it's redundant with mbedtls_cipher_info_from_psa which enumerates
  supported key types explicitly.
* The order of some validations is different to allow the split between
  setup and data processing. The code now calls a more robust function
  psa_aead_abort in case of any error after the early stage of the setup.
2018-10-08 14:42:11 +02:00
Gilles Peskine
d911eb7915 Add support for truncated MAC algorithms 2018-10-08 14:42:11 +02:00
Gilles Peskine
159ebf3b8e
Merge pull request #142 from ARMmbed/psa-metadata_validation
Algorithm and key type encoding validation
2018-09-24 10:38:39 +02:00
Jaeden Amero
3bfe5f0388
Merge pull request #150 from ARMmbed/psa-rsa-verify-error
Don't return INVALID_PADDING on an invalid RSA signature
2018-09-21 16:15:10 +01:00
Jaeden Amero
40363fe3f5
Merge pull request #139 from ARMmbed/psa-PSA_ALG-block_cipher_padding
Correct and simplify block-based cipher modes
2018-09-21 09:38:27 +01:00
Gilles Peskine
00709fafb8 Rename PSA_ALG_HMAC_HASH to PSA_ALG_HMAC_GET_HASH
Be consistent with other GET_HASH macros.
2018-09-20 12:00:21 +02:00
Gilles Peskine
daea26f70f Correct and simplify block-based cipher modes
OFB and CFB are streaming modes. XTS is a not a cipher mode but it
doesn't use a separate padding step. This leaves only CBC as a block
cipher mode that needs a padding step.

Since CBC is the only mode that uses a separate padding step, and is
likely to remain the only mode in the future, encode the padding mode
directly in the algorithm constant, rather than building up an
algorithm value from a chaining mode and a padding mode. This greatly
simplifies the interface as well as some parts of the implementation.
2018-09-20 11:58:22 +02:00
Gilles Peskine
ef12c63de0 RSA verification: don't report an invalid padding error
Mbed TLS distinguishes "invalid padding" from "valid padding but the
rest of the signature is invalid". This has little use in practice and
PSA doesn't report this distinction. We just report "invalid
signature".
2018-09-17 14:33:43 +02:00
itayzafrir
90d8c7a728 Ensure the module is initialized in key based functions 2018-09-16 12:15:47 +03:00
itayzafrir
0adf0fc31c Ensure the module is initialized in psa_generate_random 2018-09-16 12:15:46 +03:00
Gilles Peskine
78b3bb670d Change the bitwise encoding of key type categories
There were only 5 categories (now 4). Reduce the category mask from 7
bits to 3.

Combine unformatted, not-necessarily-uniform keys (HMAC, derivation)
with raw data.

Reintroduce a KEY_TYPE_IS_UNSTRUCTURED macro (which used to exist
under the name KEY_TYPE_IS_RAW_DATA macro) for key types that don't
have any structure, including both should-be-uniform keys (such as
block cipher and stream cipher keys) and not-necessarily-uniform
keys (such as HMAC keys and secrets for key derivation).
2018-09-14 10:02:29 +01:00
itayzafrir
534bd7c33b Add missing calls to psa_cipher_abort in cipher functions 2018-09-14 10:02:29 +01:00
itayzafrir
40835d4e56 Add missing calls to psa_hash_abort in hash functions 2018-09-14 10:02:29 +01:00
Jaeden Amero
df3b74f168 Merge pull request #125 from ARMmbed/key_slot_index-fix
Fix off-by-one errors in key slot index limits
2018-09-14 10:02:29 +01:00
Gilles Peskine
9a05634558 psa_crypto_free: destroy the last slot
The last slot in the array was not freed due to an off-by-one error.

Amend the fill_slots test to serve as a non-regression test for this
issue: without this bug fix, it would cause a memory leak.
2018-09-14 10:02:29 +01:00
Gilles Peskine
996deb18cc Fix buffer overflow in the slot array
Slots are numbered from 1, but the slot array is a C array so it's
numbered from 0.

Add a non-regression test.
2018-09-14 10:02:29 +01:00
Darryl Green
9e2d7a09f1 Add ifdefs for psa_internal_export_key function
MBEDTLS_PK_WRITE_C only requires either MBEDTLS_RSA_C or MBEDTLS_ECP_C to be defined.
Added wrappers to handle the cases where only one has been defined.
Moved mbedtls_pk_init to be within the ifdefs, so it's only called if appropriate.
2018-09-14 10:02:28 +01:00
Darryl Green
8800136156 Fix key parameter in psa_key_derivation to use correct type 2018-09-14 10:02:28 +01:00
Darryl Green
8f8aa8f931 Add ifdef for MBEDTLS_ECP_C around ecc_group_to_psa functions
These are unused if MBEDTLS_ECP_C isn't defined.
2018-09-14 10:02:28 +01:00
Darryl Green
5cc689a821 Add handling for (sometimes) unused parameters 2018-09-14 10:02:28 +01:00
Gilles Peskine
9f900a8b25 Merge pull request #105 from ARMmbed/psa-derive_hkdf
PSA key derivation: simple-ish interface, HKDF
2018-09-12 16:50:07 +03:00
Gilles Peskine
08542d8037 Fix psa_generator_import_key for DES
In psa_generator_import_key, if generating a DES or 3DES key, set the
parity bits.

Add tests for deriving a DES key. Also test deriving an AES key while
I'm at it.
2018-09-12 16:50:07 +03:00
Gilles Peskine
d54931c7c4 HKDF: be more robust if we reach the maximum ouptut length
In psa_generator_hkdf_read, return BAD_STATE if we're trying to
construct more output than the algorithm allows. This can't happen
through the API due to the capacity limit, but it could potentially
happen in an internal call.

Also add a test case that verifies that we can set up HKDF with its
maximum capacity and read up to the maximum capacity.
2018-09-12 16:50:05 +03:00
Gilles Peskine
bef7f14f8e Implement HKDF 2018-09-12 16:44:05 +03:00
Gilles Peskine
ea0fb4975c Add framework for simple key derivation
New key type PSA_KEY_TYPE_DERIVE. New usage flag PSA_KEY_USAGE_DERIVE.
New function psa_key_derivation.

No key derivation algorithm is implemented yet. The code may not
compile with -Wunused.

Write some unit test code for psa_key_derivation. Most of it cannot be
used yet due to the lack of a key derivation algorithm.
2018-09-12 16:44:04 +03:00
Gilles Peskine
eab56e4159 Add generator API
Add an API for byte generators: psa_crypto_generator_t,
PSA_CRYPTO_GENERATOR_INIT, psa_crypto_generator_init,
psa_get_generator_capacity, psa_generator_read,
psa_generator_import_key, psa_generator_abort.

This commit does not yet implement any generator algorithm, it only
provides the framework. This code may not compile with -Wunused.
2018-09-12 16:41:12 +03:00
Gilles Peskine
9aa369eafb HMAC: improve robustness checks on hash/block size
In psa_mac_setup and psa_hmac_setup_internal, perform a sanity check
on the hash size and the hash block size respectively. These sanity
checks should only trigger on an incompletely or incorrectly
implemented hash function.

Remove the check on the block size in psa_hmac_finish_internal
because at this point it has already been checked and used.
2018-09-12 16:41:12 +03:00
Gilles Peskine
1e6bfdff5e psa_hmac_setup_internal: fix double call of psa_hash_setup
In the common case (key no longer than the block size), psa_hash_setup
was being called twice in succession. With current implementations
this is just a small performance loss, but potentially with
alternative implementations this could have lead to a memory leak.
2018-09-12 16:41:12 +03:00
Gilles Peskine
b3fc05d776 psa_asymmetric_{encrypt,decrypt}: reject salt when not allowed
In psa_asymmetric_encrypt and psa_asymmetric_decrypt, if the algorithm
does not use a salt, require the salt to be empty.
2018-09-12 16:41:12 +03:00
Gilles Peskine
ff94abdf3a Make psa_hmac_setup_internal more standalone
Call psa_hash_setup in psa_hmac_setup_internal rather than
psa_mac_init. This makes it easier to use psa_hmac_setup_internal on
its own (for the sake of using HMAC internally inside the library).
2018-09-12 16:41:12 +03:00
Gilles Peskine
b8be288374 psa_hmac_setup_internal: add some missing cleanup on failure
Clean ipad if hashing the key failed.
2018-09-12 16:41:12 +03:00
Gilles Peskine
9688997301 MAC setup: support 0-length HMAC key
Avoid undefined behavior when using a 0-length HMAC key (Asan
complained).
2018-09-12 16:41:12 +03:00
Gilles Peskine
072ac56a03 Implement OAEP
Implement RSAES-OAEP encryption and decryption.

Test it to the same level as PKCS#1 v1.5.
2018-09-12 16:41:12 +03:00
Gilles Peskine
01126fae7f Isolate HMAC code into its own functions
Create internal functions for HMAC operations. This prepares for two
things: separating crypto-sensitive code from argument decoding and
validation, and using HMAC for other purposes than a MAC inside the
library (e.g. HMAC_DRBG, HKDF).

No intended observable behavior change in this commit.
2018-09-12 16:41:12 +03:00
Gilles Peskine
71ac7b11a7 Allow RSA PSS with any input size
Although RSASSA-PSS defines its input as a message to be hashed, we
implement a sign-the-hash function. This function can take an input
which isn't a hash, so don't restrict the size of the input, any more
than Mbed TLS does.

Remove a redundant check that hash_length fits in unsigned int for the
sake of Mbed TLS RSA functions.

Test that PSS accepts inputs of various lengths. For PKCS#1 v1.5
signature in raw mode, test the maximum input length.
2018-09-12 16:41:12 +03:00
Gilles Peskine
94e44540ff psa_hash_update: robustify the case length=0
Don't require hash implementations to behave correctly on a
zero-length input, which may have an invalid pointer.
2018-09-12 16:41:12 +03:00
Gilles Peskine
af89fd771e psa_import_key: split RSA and ECP code into small functions
No intended behavior change except that some edge cases may have
swapped between INVALID_ARGUMENT and NOT_SUPPORTED.
2018-09-12 16:41:11 +03:00