Commit graph

9034 commits

Author SHA1 Message Date
Gilles Peskine f3b731e817 Move integral types and associated macros to their own header
Some parts of the library, and crypto drivers, need to see key types,
algorithms, policies, etc. but not API functions. Move portable
integral types and macros to build and analyze values of these types
to a separate headers crypto_types.h and crypto_values.h.

No functional changes, code was only moved from crypto.h to the new headers.
2018-12-21 17:53:09 +01:00
Gilles Peskine 5e09bc7eb5 Fix maybe-uninitialized warning
GCC 4.8 warns that some variables may be used without having been
initialized. They aren't, but determining that takes nontrivial
analysis, so initialize them at the point of definition.
2018-12-21 12:06:15 +01:00
Gilles Peskine 0344d8171d Simplify the SPM compatibility hack
Define psa_status_t to int32_t unconditionally. There's no reason to
refer to psa_error_t here: psa_error_t is int32_t if it's present. We
would only need a conditional definition if psa_defs.h and
psa_crypto.h used the same type name.

Keep the conditional definition of PSA_SUCCESS. Although the C
preprocessor allows a duplicate definition for a macro, it has to be
the exact same token sequence, not merely an equivalent way to build
the same value.
2018-12-20 20:09:04 +01:00
Gilles Peskine c08fc1d7e9 Move MIN and MAX macros from PSA tests to helpers.function 2018-12-20 18:47:53 +01:00
Gilles Peskine d76f181617 Prefer ASSERT_ALLOC to calloc+TEST_ASSERT in PSA tests
To allocate memory dynamically in a test, call ASSERT_ALLOC which
takes care of calling calloc and of checking for NULL.
2018-12-20 18:47:53 +01:00
Gilles Peskine 1f2aa0e3b0 Remove useless null checks of data_t* parameters
The test framework never passes NULL for a data_t* parameter, so
testing them against NULL is clutter.
2018-12-20 18:47:53 +01:00
Gilles Peskine 40ab95bdbc Remove checks of test parameters against SIZE_MAX
Our code base doesn't even support 16-bit platforms, so those checks
are always trivially true.
2018-12-20 18:47:53 +01:00
Gilles Peskine 0dfba2ddf0 Use ASSERT_COMPARE in preference to memcmp in PSA tests 2018-12-20 18:47:52 +01:00
Gilles Peskine f812dcf4ae 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.
2018-12-20 18:47:52 +01:00
Gilles Peskine fe11b72b93 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
2018-12-20 18:47:52 +01:00
Gilles Peskine 8817f61007 Use PSA_ASSERT(a) in preference to TEST_ASSERT(a==PSA_SUCCESS)
This commit is the result of the following command, followed by
reindenting (but not wrapping lines):

perl -00 -i -pe 's/^( *)TEST_ASSERT\(([^;=]*)(?: |\n *)==\s*PSA_SUCCESS\s*\);$/${1}PSA_ASSERT($2 );/gm' tests/suites/test_suite_psa_*.function
2018-12-20 18:47:52 +01:00
Gilles Peskine 0f915f1d2a Indent PSA tests according to K&R rules with Mbed TLS tweaks
Only whitespace changes in this commit.
2018-12-20 18:47:52 +01:00
Gilles Peskine 9d8eea7e19 Wrap some multiline expressions in parentheses
This guarantees that they'll be indented as desired under most
indentation rules.
2018-12-20 18:47:52 +01:00
Gilles Peskine 5f7aeeea06 New test macro TEST_EQUAL
TEST_EQUAL(expr1, expr2) is just TEST_ASSERT((expr1) == (expr2)) for
now, but in the future I hope that it will print out the differing
values.
2018-12-20 18:47:52 +01:00
Gilles Peskine 0174be2c17 Move the PSA_ASSERT macro to the common helpers file
It's potentially useful in all PSA test suites, of which there are now
several.
2018-12-20 18:47:52 +01:00
Gilles Peskine f055ad7512 Add a safety check to ARRAY_LENGTH
Cause a compilation error on ARRAY_LENGTH(p) where p is a pointer as
opposed to an array. This only works under GCC and compatible
compilers such as Clang. On other compilers, ARRAY_LENGTH works but
doesn't check the type of its argument.
2018-12-20 18:47:52 +01:00
Gilles Peskine 3d2f949c86 Move the ARRAY_LENGTH macro to the common helpers file 2018-12-17 23:17:17 +01:00
Gilles Peskine 1a76f3971c Merge remote-tracking branch 'psa/pr/230' into feature-psa
Implement a key handle mechanism
2018-12-12 16:15:34 +01:00
Gilles Peskine 0a0bd7b3d1
Merge pull request #239 from ARMmbed/feature-psa-no_test_USE_PSA_CRYPTO
Remove testing of MBEDTLS_USE_PSA_CRYPTO
2018-12-12 14:07:05 +01:00
Gilles Peskine 2e14bd3aaf Add missing static on file-scope variable 2018-12-12 14:05:18 +01:00
Gilles Peskine 76965bb678 Remove MBEDTLS_USE_PSA_CRYPTO tests from all.sh
This branch isn't expected to keep USE_PSA_CRYPTO working.
2018-12-12 10:48:06 +01:00
Gilles Peskine 23fd2bdb94 Update some documentation related to key slots
Some of the documentation is obsolete in its reference to key slots
when it should discuss key handles. This may require a further pass,
possibly with some reorganization of error codes.

Update the documentation of functions that modify key slots (key
material creation and psa_set_key_policy()) to discuss how they affect
storage.
2018-12-11 16:48:14 +01:00
Gilles Peskine 79a11d6c42 Test invalid lifetime regardless of support for persistent keys
Even if persistent keys are not supported, psa_open_key and
psa_create_key must fail if invoked with an invalid lifetime value.
2018-12-11 16:48:14 +01:00
Gilles Peskine 09829036ab Document some functions in internal headers 2018-12-11 16:48:14 +01:00
Gilles Peskine fa4135b135 Move more slot management functions to the proper module
Move psa_load_persistent_key_into_slot,
psa_internal_make_key_persistent and psa_internal_release_key_slot to
the slot management module.

Expose psa_import_key_into_slot from the core.

After this commit, there are no longer any functions declared in
psa_crypto_slot_management.h and defined in psa_crypto.c. There are
still function calls in both directions between psa_crypto.c and
psa_crypto_slot_management.c.
2018-12-11 16:48:14 +01:00
Gilles Peskine 66fb126e87 Move the key slot array to the slot management module
Move the key slot array and its initialization and wiping to the slot
management module.

Also move the lowest-level key slot access function psa_get_key_slot
and the auxiliary function for slot allocation
psa_internal_allocate_key_slot to the slot management module.
2018-12-11 16:48:14 +01:00
Gilles Peskine 408319be3a Remove redundant check for slot->allocated
This check became redundant when support for direct access to key
slots was removed.
2018-12-11 16:48:14 +01:00
Gilles Peskine 539cda57df Add tests with invalid key identifiers
Test that 0 and PSA_CRYPTO_ITS_RANDOM_SEED_UID are not accepted as key
identifiers.
2018-12-11 16:48:14 +01:00
Gilles Peskine 48868129cd Document the maximum key identifier
Give it a name and explain why it was chosen.
2018-12-11 16:48:14 +01:00
Gilles Peskine 039b90cf5b Expose the PSA key slot structure to internal modules
Move psa_key_slot_t to a new header psa_crypto_core.h, to prepare for
moving the responsibility for some fields to
psa_crypto_slot_management.c.
2018-12-11 16:48:14 +01:00
Gilles Peskine 3f10812ff9 Use the library-wide zeroize function
Since Mbed TLS 2.10, there is a single copy of
mbedtls_platform_zeroize for the whole library instead of one per
module. Update the PSA crypto module accordingly.
2018-12-11 16:48:14 +01:00
Gilles Peskine fe9756b1f5 Remove a comment that is no longer applicable 2018-12-11 16:48:14 +01:00
Gilles Peskine 2f060a8ea5 Rename key_slot_t to psa_key_slot_t in psa_crypto.c
That way it'll be ok to add it to an internal header so as to use it
in multiple source modules.
2018-12-11 16:48:14 +01:00
Gilles Peskine 4a044739a8 Fix the build without persistent storage
Add missing guards on MBEDTLS_PSA_CRYPTO_STORAGE_C.

Add test cases to test that psa_create_key and psa_open_key return
NOT_SUPPORTED.
2018-12-11 16:48:14 +01:00
Gilles Peskine c5487a8892 Drop support for psa_key_slot_t in psa_crypto.c
This commit finishes the removal of support for direct access to key
slots in psa_crypto.c.

This marks the end of the necessary phase of the transition to key
handles. The code should subsequently be refactored to move key slot
management from psa_crypto.c to psa_crypto_slot_management.c.
2018-12-11 16:48:13 +01:00
Gilles Peskine b77a6b25c0 Remove psa_set_key_lifetime
This function is no longer relevant. Use psa_create_key instead.
2018-12-11 16:48:13 +01:00
Gilles Peskine f6cc435a8a Remove psa_key_slot_t from public headers
This commit marks the beginning of the removal of support for direct
access to key slots. From this commit on, programs that use
psa_key_slot_t will no longer compile.

Subsequent commits will remove the now-unused legacy support in
psa_crypto.c.
2018-12-11 16:48:13 +01:00
Gilles Peskine 7bc9f68232 Convert the PSA crypto persistent storage tests to the new handle API
Switch from the direct use of slot numbers to handles allocated by
psa_allocate_key.

The general principle for each function is:
* Change `psa_key_slot_t slot` to `psa_key_handle_t handle` or
  `psa_key_id_t key_id` depending on whether it's used as a handle to
  an open slot or as a persistent name for a key.
* Call psa_create_key() before using a slot, instead of calling
  psa_set_key_lifetime to make a slot persistent.

Remove the unit test persistent_key_is_configurable which is no longer
relevant.
2018-12-11 16:48:13 +01:00
Gilles Peskine a23eafce3a Fix snprintf call to assume less about integral type sizes
The code only worked if psa_key_id_t (formerly psa_key_slot_t)
promoted to int and every value fit in int. Now the code only assumes
that psa_key_id_t is less wide than unsigned long, which is the case
since psa_key_id_t is a 32-bit type in our implementation.
2018-12-11 16:48:13 +01:00
Gilles Peskine 8d4919bc6a Persistent storage implementation: psa_key_slot_t -> psa_key_id_t
Move the persistent storage implementation from psa_key_slot_t to
psa_key_id_t. For the most part, this just means changing the types of
function arguments.

Update the documentation of some functions to reflect the fact that
the slot identifier is purely a storage identifier and is not related
to how the slot is designated in memory.
2018-12-11 16:48:13 +01:00
Gilles Peskine b0edfb513b Convert the PSA example programs to the new handle API
Switch from the direct use of slot numbers to handles allocated by
psa_allocate_key.
2018-12-11 16:48:13 +01:00
Gilles Peskine bdf309ccdb Convert the PSA crypto cryptography tests to the new handle API
Switch from the direct use of slot numbers to handles allocated by
psa_allocate_key.

This commit does not affect persistent key tests except for the one
test function in test_suite_psa_crypto that uses persistent keys
(persistent_key_load_key_from_storage).

The general principle for each function is:
* Change `psa_key_slot_t slot` to `psa_key_handle_t handle`.
* Call psa_allocate_key() before setting the policy of the slot,
  or before creating key material in functions that don't set a policy.
* Some PSA_ERROR_EMPTY_SLOT errors become PSA_ERROR_INVALID_HANDLE
  because there is now a distinction between not having a valid
  handle, and having a valid handle to a slot that doesn't contain key
  material.
* In tests that use symmetric keys, calculate the max_bits parameters
  of psa_allocate_key() from the key data size. In tests where the key
  may be asymmetric, call an auxiliary macro KEY_BITS_FROM_DATA which
  returns an overapproximation. There's no good way to find a good
  value for max_bits with the API, I think the API should be tweaked.
2018-12-11 16:48:13 +01:00
Gilles Peskine a8860b2990 Remove lifetime test functions
With the handle-based slot management interface, psa_set_key_lifetime
will no longer exist, so remove the corresponding unit tests.
2018-12-11 16:48:13 +01:00
Gilles Peskine dc911fd594 Remove redundant slot-based test
fill_slots is superseded by many_transient_handles.
2018-12-11 16:48:13 +01:00
Gilles Peskine f77ed1f20b Factor the common idiom psa_wipe_key_slot into a function
Many places in the code called psa_remove_key_data_from_memory (which
preserves metadata for the sake of failues in psa_import_key) followed
by clearing the slot data. Use an auxiliary function for this.
2018-12-11 16:48:13 +01:00
Gilles Peskine a426168cbf Test that failure of import_key preserves metadata 2018-12-11 16:48:13 +01:00
Gilles Peskine d7c75707b8 mbedtls_psa_crypto_free: free allocated slots as well
Access the slot directly rather than going through psa_get_key_slot.
Unlike other places where key slots are accessed through
psa_get_key_slot, here, we know where all the slots are and there are
no policy or permission considerations.

This resolves a memory leak: allocated slots were not getting freed
because psa_get_key_slot rejected the attempt of accessing them
directly rather than via a handle.
2018-12-11 16:48:13 +01:00
Gilles Peskine 961849f6d1 Implement slot allocation
Implement psa_allocate_key, psa_open_key, psa_create_key,
psa_close_key.

Add support for keys designated to handles to psa_get_key_slot, and
thereby to the whole API.

Allocated and non-allocated keys can coexist. This is a temporary
stage in order to transition from the use of direct slot numbers to
allocated handles only. Once all the tests and sample programs have
been migrated to use handles, the implementation will be simplified
and made more robust with support for handles only.
2018-12-11 16:48:13 +01:00
Gilles Peskine 5ec7b078ea Add tests for the new slot management mechanism
Add unit tests for handle allocation and release.
2018-12-11 16:48:12 +01:00
Gilles Peskine 7f6e3a868a Change ASSERT_ALLOC to take a size in elements, not bytes
`ASSERT_ALLOC(p, length)` now allocates `length` elements, i.e.
`length * sizeof(*p)` bytes.
2018-12-11 16:48:12 +01:00