The function mbedtls_mpi_sub_abs first checked that A >= B and then
performed the subtraction, relying on the fact that A >= B to
guarantee that the carry propagation would stop, and not taking
advantage of the fact that the carry when subtracting two numbers can
only be 0 or 1. This made the carry propagation code a little hard to
follow.
Write an ad hoc loop for the carry propagation, checking the size of
the result. This makes termination obvious.
The initial check that A >= B is no longer needed, since the function
now checks that the carry propagation terminates, which is equivalent.
This is a slight performance gain.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
There was some confusion during review about when A->p[n] could be
nonzero. In fact, there is no need to set A->p[n]: only the
intermediate result d might need to extend to n+1 limbs, not the final
result A. So never access A->p[n]. Rework the explanation of the
calculation in a way that should be easier to follow.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
The function mpi_sub_hlp had confusing semantics: although it took a
size parameter, it accessed the limb array d beyond this size, to
propagate the carry. This made the function difficult to understand
and analyze, with a potential buffer overflow if misused (not enough
room to propagate the carry).
Change the function so that it only performs the subtraction within
the specified number of limbs, and returns the carry.
Move the carry propagation out of mpi_sub_hlp and into its caller
mbedtls_mpi_sub_abs. This makes the code of subtraction very slightly
less neat, but not significantly different.
In the one other place where mpi_sub_hlp is used, namely mpi_montmul,
this is a net win because the carry is potentially sensitive data and
the function carefully arranges to not have to propagate it.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
mpi_sub_hlp performs a subtraction A - B, but took parameters in the
order (B, A). Swap the parameters so that they match the usual
mathematical syntax.
This has the additional benefit of putting the output parameter (A)
first, which is the normal convention in this module.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Checking the budget only after the randomization is done means sometimes we
were randomizing first, then noticing we ran out of budget, return, come back
and randomize again before we finally normalize.
While this is fine from a correctness and security perspective, it's a minor
inefficiency, and can also be disconcerting while debugging, so we might as
well avoid it.
Signed-off-by: Manuel Pégourié-Gonnard <manuel.pegourie-gonnard@arm.com>
It results in smaller code than using CTR_DRBG (64 bytes smaller on ARMv6-M
with arm-none-eabi-gcc 7.3.1), so let's use this by default when both are
available.
Signed-off-by: Manuel Pégourié-Gonnard <manuel.pegourie-gonnard@arm.com>
Unless MBEDTLS_ECP_NO_INTERNAL_RNG is defined, it's no longer possible for
f_rng to be NULL at the places that randomize coordinates.
Eliminate the NULL check in this case:
- it makes it clearer to reviewers that randomization always happens (unless
the user opted out at compile time)
- a NULL check in a place where it's easy to prove the value is never NULL
might upset or confuse static analyzers (including humans)
- removing the check saves a bit of code size
Signed-off-by: Manuel Pégourié-Gonnard <manuel.pegourie-gonnard@arm.com>
Currently we draw pseudo-random numbers at the beginning and end of the main
loop. With ECP_RESTARTABLE, it's possible that between those two occasions we
returned from the multiplication function, hence lost our internal DRBG
context that lives in this function's stack frame. This would result in the
same pseudo-random numbers being used for blinding in multiple places. While
it's not immediately clear that this would give rise to an attack, it's also
absolutely not clear that it doesn't. So let's avoid that by using a DRBG
context that lives inside the restart context and persists across
return/resume cycles. That way the RESTARTABLE case uses exactly the
same pseudo-random numbers as the non-restartable case.
Testing and compile-time options:
- The case ECP_RESTARTABLE && !ECP_NO_INTERNAL_RNG is already tested by
component_test_no_use_psa_crypto_full_cmake_asan.
- The case ECP_RESTARTABLE && ECP_NO_INTERNAL_RNG didn't have a pre-existing
test so a component is added.
Testing and runtime options: when ECP_RESTARTABLE is enabled, the test suites
already contain cases where restart happens and cases where it doesn't
(because the operation is short enough or because restart is disabled (NULL
restart context)).
Signed-off-by: Manuel Pégourié-Gonnard <manuel.pegourie-gonnard@arm.com>
While it seems cleaner and more convenient to set it in the top-level
mbedtls_ecp_mul() function, the existence of the restartable option changes
things - when it's enabled the drbg context needs to be saved in the restart
context (more precisely in the restart_mul sub-context), which can only be
done when it's allocated, which is in the curve-specific mul function.
This commit only internal drbg management from mbedtls_ecp_mul() to
ecp_mul_mxz() and ecp_mul_comb(), without modifying behaviour (even internal),
and a future commit will modify the ecp_mul_comb() version to handle restart
properly.
Signed-off-by: Manuel Pégourié-Gonnard <manuel.pegourie-gonnard@arm.com>
The case of MBEDTLS_ECP_RESTARTABLE isn't handled correctly yet: in that case
the DRBG instance should persist when resuming the operation. This will be
addressed in the next commit.
When both CTR_DRBG and HMAC_DRBG are available, CTR_DRBG is preferred since
both are suitable but CTR_DRBG tends to be faster and I needed a tie-breaker.
There are currently three possible cases to test:
- NO_INTERNAL_RNG is set -> tested in test_ecp_no_internal_rng
- it's unset and CTR_DRBG is available -> tested in the default config
- it's unset and CTR_DRBG is disabled -> tested in
test_ecp_internal_rng_no_ctr_drbg
Signed-off-by: Manuel Pégourié-Gonnard <manuel.pegourie-gonnard@arm.com>
No effect so far, except on dependency checking, as the feature it's meant to
disable isn't implemented yet (so the descriptions in config.h and the
ChangeLog entry are anticipation for now).
Signed-off-by: Manuel Pégourié-Gonnard <manuel.pegourie-gonnard@arm.com>
Let code analyzers know that this is deliberate. For example MSVC
warns about the conversion if it's implicit.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
In mpi_montmul, an auxiliary function for modular
exponentiation (mbedtls_mpi_mod_exp) that performs Montgomery
multiplication, the last step is a conditional subtraction to force
the result into the correct range. The current implementation uses a
branch and therefore may leak information about secret data to an
adversary who can observe what branch is taken through a side channel.
Avoid this potential leak by always doing the same subtraction and
doing a contant-trace conditional assignment to set the result.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Separate out a version of mpi_safe_cond_assign that works on
equal-sized limb arrays, without worrying about allocation sizes or
signs.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
This reverts commit 2cc69fffcf.
A check was added in mpi_montmul because clang-analyzer warned about a
possibly null pointer. However this was a false positive. Recent
versions of clang-analyzer no longer emit a warning (3.6 does, 6
doesn't).
Incidentally, the size check was wrong: mpi_montmul needs
T->n >= 2 * (N->n + 1), not just T->n >= N->n + 1.
Given that this is an internal function which is only used from one
public function and in a tightly controlled way, remove both the null
check (which is of low value to begin with) and the size check (which
would be slightly more valuable, but was wrong anyway). This allows
the function not to need to return an error, which makes the source
code a little easier to read and makes the object code a little
smaller.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
The previous version attempted to write the explicit IV from
the destination buffer before it has been written there.
Signed-off-by: Hanno Becker <hanno.becker@arm.com>
Invasive testing strategy
Create a new header `common.h`.
Introduce a configuration option `MBEDTLS_TEST_HOOKS` for test-specific code, to be used in accordance with the invasive testing strategy.
This is to avoid confusion with the class of macros
MBEDTLS_SSL_PROTO_TLS1_X
which have an underscore between major and minor version number.
Signed-off-by: Hanno Becker <hanno.becker@arm.com>
TLS 1.3 record protection allows the addition of an arbitrary amount
of padding.
This commit introduces a configuration option
```
MBEDTLS_SSL_TLS13_PADDING_GRANULARITY
```
The semantics of this option is that padding is chosen in a minimal
way so that the padded plaintext has a length which is a multiple of
MBEDTLS_SSL_TLS13_PADDING_GRANULARITY.
For example, setting MBEDTLS_SSL_TLS13_PADDING_GRANULARITY to 1024
means that padded plaintexts will have length 1024, 2048, ..., while
setting it to 1 means that no padding will be used.
Signed-off-by: Hanno Becker <hanno.becker@arm.com>
This commit adds an error condition for bad user configurations
and updates the number of SSL module errors in error.h.
Signed-off-by: Ronald Cron <ronald.cron@arm.com>
This commit uses the previously defined macro to uniformize
bounds checks in several places. It also adds bounds checks to
the ClientHello writing function that were previously missing.
Also, the functions adding extensions to the ClientHello message
can now fail if the buffer is too small or a different error
condition occurs, and moreover they take an additional buffer
end parameter to free them from the assumption that one is
writing to the default output buffer.
Signed-off-by: Ronald Cron <ronald.cron@arm.com>
new name: mbedtls_x509_crt_parse_der_with_ext_cb
Co-authored-by: Gilles Peskine <gilles.peskine@arm.com>
Signed-off-by: Nicola Di Lieto <nicola.dilieto@gmail.com>
The structure `mbedtls_ssl_transform` representing record protection
transformations should ideally be used through a function-based
interface only, as this will ease change of implementation as well
as the addition of new record protection routines in the future.
This commit makes a step in that direction by introducing the
helper function `ssl_transform_get_explicit_iv_len()` which
returns the size of the pre-expansion during record encryption
due to the potential addition of an explicit IV.
Signed-off-by: Hanno Becker <hanno.becker@arm.com>
This commit simplifies nonce derivation for AEAD based record protection
routines in the following way.
So far, code distinguished between the cases of GCM+CCM and ChachaPoly:
- In the case of GCM+CCM, the AEAD nonce is the concatentation
of a 4-byte Fixed IV and a dynamically chosen 8-byte IV which is prepended
to the record. In Mbed TLS, this is always chosen to be the record sequence
number, but it need not to.
- In the case of ChaChaPoly, the AEAD nonce is derived as
`( 12-byte Fixed IV ) XOR ( 0 || 8-byte dynamic IV == record seq nr )`
and the dynamically chosen IV is no longer prepended to the record.
This commit removes this distinction by always computing the record nonce
via the formula
`IV == ( Fixed IV || 0 ) XOR ( 0 || Dynamic IV )`
The ChaChaPoly case is recovered in case `Len(Fixed IV) == Len(IV)`, and
GCM+CCM is recovered when `Len(IV) == Len(Fixed IV) + Len(Dynamic IV)`.
Moreover, a getter stub `ssl_transform_aead_dynamic_iv_is_explicit()`
is introduced which infers from a transform whether the dynamically
chosen part of the IV is explicit, which in the current implementation
of `mbedtls_ssl_transform` can be derived from the helper field
`mbedtls_ssl_transform::fixed_ivlen`.
Signed-off-by: Hanno Becker <hanno.becker@arm.com>
The computation of the per-record nonce for AEAD record protection
varies with the AEAD algorithm and the TLS version in use.
This commit introduces a helper function for the nonce computation
to ease readability of the quite monolithic record encrytion routine.
Signed-off-by: Hanno Becker <hanno.becker@arm.com>
The previous record protection code added the explicit part of the
record nonce prior to encrypting the record. This temporarily leaves
the record structure in the undesireable state that the data outsie
of the interval `rec->data_offset, .., rec->data_offset + rec->data_len`
has already been written.
This commit moves the addition of the explicit IV past record encryption.
Signed-off-by: Hanno Becker <hanno.becker@arm.com>
The internal functions
`ssl_cid_{build/parse}_inner_plaintext()`
implement the TLSInnerPlaintext mechanism used by DTLS 1.2 + CID
in order to allow for flexible length padding and to protect the
true content type of a record.
This feature is also present in TLS 1.3 support for which is under
development. As a preparatory step towards sharing the code between
the case of DTLS 1.2 + CID and TLS 1.3, this commit renames
`ssl_cid_{build/parse}_inner_plaintext()`
to
`ssl_{build/parse}_inner_plaintext()`.
Signed-off-by: Hanno Becker <hanno.becker@arm.com>
Now that lifetimes have structures and secure element drivers handle
all the lifetimes with a certain location, update driver registration
to take a location as argument rather than a lifetime.
This commit updates the Mbed TLS implementation.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
The ssl_cli.c:ssl_write_supported_elliptic_curves_ext()
function is compiled only if MBEDTLS_ECDH_C, MBEDTLS_ECDSA_C
or MBEDTLS_KEY_EXCHANGE_ECJPAKE_ENABLED is defined which
implies that MBEDTLS_ECP_C is defined. Thus remove the
precompiler conditions on MBEDTLS_ECP_C in its code.
Signed-off-by: Ronald Cron <ronald.cron@arm.com>
On dual world platforms, we want to run the PK module (pk.c) on the NS
side so TLS can use PSA APIs via the PK interface. PK currently has a
hard dependency on mbedtls_ecc_group_to_psa() which is declared in
crypto_extra.h, but only defined in psa_crypto.c, which is only built
for the S side.
Without this change, dual world platforms get error messages like the
following.
[Error] @0,0: L6218E: Undefined symbol mbedtls_ecc_group_to_psa (referred from BUILD/LPC55S69_NS/ARM/mbed-os/features/mbedtls/mbed-crypto/src/pk.o)
Make mbedtls_ecc_group_to_psa() inline within crypto_extra.h so that it
is available to both NS and S world code.
Fixes#3300
Signed-off-by: Darryl Green <darryl.green@arm.com>
Signed-off-by: Jaeden Amero <jaeden.amero@arm.com>
Clear bits in mbedtls_ecc_group_to_psa() to avoid static analyzers and
possibly compilers from warning that bits may be used uninitialized in
certain code paths.
For example, if mbedtls_ecc_group_to_psa() were to be inlined in
crypto_extra.h, the following compiler warning is likely.
In file included from ../include/psa/crypto.h:3774:0,
from ../include/mbedtls/pk.h:49,
from pk.c:29:
pk.c: In function 'mbedtls_pk_wrap_as_opaque':
../include/psa/crypto_struct.h:460:33: error: 'bits' may be used uninitialized in this function [-Werror=maybe-uninitialized]
attributes->core.bits = (psa_key_bits_t) bits;
^~~~~~~~~~~~~~~~~~~~~
pk.c:608:12: note: 'bits' was declared here
size_t bits;
^~~~
Signed-off-by: Jaeden Amero <jaeden.amero@arm.com>
This routine is functionally equivalent to mbedtls_x509_crt_parse_der(),
but it accepts an additional callback function which it calls with
every unsupported certificate extension.
Proposed solution to https://github.com/ARMmbed/mbedtls/issues/3241
Signed-off-by: Nicola Di Lieto <nicola.dilieto@gmail.com>
getaddrinfo() is not available on win2k. By including wspiapi.h (if
_WIN32_WINNT is defined as value < 0x0501) then a compatibility layer
will be used when running on win2k. For more details, refer to Microsoft
docs for getaddrinfo().
Signed-off-by: opatomic <j@opatomic.com>
ecp_double_add_mxz wrongly does an MPI addition followed by a call to
MOD_MUL instead of MOD_ADD. This is more visible since the
mbedtls_mpi_xxx_mod functions have been added in commit 3b3b34f608
("Replace some macros by functions").
Fix that by using mbedtls_mpi_add_mod instead. The testsuite still
passes after that change.
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
It is sufficient to check for the strongest limit only. Using a smaller
type ensures there is no overflow (assuming size_t is at least 32 bits).
Fixes#2916
Signed-off-by: irwir <irwir@users.noreply.github.com>
1. The functions mbedtls_high_level_strerr and mbedtls_low_level_strerr
accept any error code and extract the high-level and low-level parts
respectively.
2. Documentation updates.
Signed-off-by: Gaurav Aggarwal <aggarg@amazon.com>
GCC and Clang accept
```
typedef struct foo foo_t;
typedef struct foo { ... } foo_t;
```
But this is not valid ISO C due to the redefinition of `foo_t`.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
