New name MBEDTLS_ERR_SSL_BAD_CERTIFICATE
Also, replace some instances of MBEDTLS_ERR_SSL_BAD_HS_CERTIFICATE
by MBEDTLS_ERR_SSL_DECODE_ERROR and MBEDTLS_ERR_SSL_ILLEGAL_PARAMETER
as fit.
Signed-off-by: Hanno Becker <hanno.becker@arm.com>
Also remove preprocessor logic for MBEDTLS_CONFIG_FILE, since
build_info.h alreadyy handles it.
This commit was generated using the following script:
# ========================
#!/bin/sh
git ls-files | grep -v '^include/mbedtls/build_info\.h$' | xargs sed -b -E -i '
/^#if !?defined\(MBEDTLS_CONFIG_FILE\)/i#include "mbedtls/build_info.h"
//,/^#endif/d
'
# ========================
Signed-off-by: Bence Szépkúti <bence.szepkuti@arm.com>
These macros were moved into a header and now check-names.sh is failing.
Add an MBEDTL_ prefix to the macro names to make it pass.
Signed-off-by: Janos Follath <janos.follath@arm.com>
The error is currently never returned to any function that PSA calls,
but keep mbedtls_to_psa_error up to date in case this changes.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
The functions mbedtls_pk_sign(), mbedtls_pk_sign_restartable(),
mbedtls_ecdsa_write_signature() and mbedtls_ecdsa_write_signature_restartable()
now take an extra parameter indicating the size of the output buffer for the
signature.
No change to RSA because for RSA, the output size is trivial to calculate.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
We were already rejecting them at the end, due to the fact that with the
usual (x, z) formulas they lead to the result (0, 0) so when we want to
normalize at the end, trying to compute the modular inverse of z will
give an error.
If we wanted to support those points, we'd a special case in
ecp_normalize_mxz(). But it's actually permitted by all sources (RFC
7748 say we MAY reject 0 as a result) and recommended by some to reject
those points (either to ensure contributory behaviour, or to protect
against timing attack when the underlying field arithmetic is not
constant-time).
Since our field arithmetic is indeed not constant-time, let's reject
those points before they get mixed with sensitive data (in
ecp_mul_mxz()), in order to avoid exploitable leaks caused by the
special cases they would trigger. (See the "May the Fourth" paper
https://eprint.iacr.org/2017/806.pdf)
Signed-off-by: Janos Follath <janos.follath@arm.com>
Without this parameter, it would be hard for callers to know how many bytes
of output the function wrote into the output buffer. It would be possible,
since the cumulated output must have the same length as the cumulated input,
but it would be cumbersome for the caller to keep track.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
The persistence level PSA_KEY_PERSISTENCE_READ_ONLY can now only be used
as intended, for keys that cannot be modified through normal use of the API.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
mbedtls_psa_get_stats() was written back before lifetimes were
structured as persistence and location. Fix its classification of
volatile external keys and internal keys with a non-default
persistence.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Clang was complaining and check-names.sh too
This only duplicates macros, so no impact on code size. In 3.0 we can
probably avoid the duplication by using an internal header under
library/ but this won't work for 2.16.
Signed-off-by: Manuel Pégourié-Gonnard <manuel.pegourie-gonnard@arm.com>
We were already rejecting them at the end, due to the fact that with the
usual (x, z) formulas they lead to the result (0, 0) so when we want to
normalize at the end, trying to compute the modular inverse of z will
give an error.
If we wanted to support those points, we'd a special case in
ecp_normalize_mxz(). But it's actually permitted by all sources
(RFC 7748 say we MAY reject 0 as a result) and recommended by some to
reject those points (either to ensure contributory behaviour, or to
protect against timing attack when the underlying field arithmetic is
not constant-time).
Since our field arithmetic is indeed not constant-time, let's reject
those points before they get mixed with sensitive data (in
ecp_mul_mxz()), in order to avoid exploitable leaks caused by the
special cases they would trigger. (See the "May the Fourth" paper
https://eprint.iacr.org/2017/806.pdf)
Signed-off-by: Manuel Pégourié-Gonnard <manuel.pegourie-gonnard@arm.com>
* development: (66 commits)
Document the return type change in the migration guide
Fix return type of example key export callbacks
Add change log
psa: mac: Add driver dispatch tests for psa_mac_verify
psa: mac: Add driver delegation support for psa_mac_verify()
psa: mac: Introduce psa_mac_compute_internal
psa: mac: Add driver dispatch tests for psa_mac_compute
psa: mac: Improve MAC finalization code
psa: mac: Add driver delegation support for psa_mac_compute()
psa: mac: Add MAC compute builtin implementation
psa: mac: Improve implementation of psa_mac_finalize_alg_and_key_validation()
psa: mac: Split psa_mac_setup()
psa: mac: Re-organize psa_mac_setup() internal function
Move export callback and context to the end of SSL context
Improve ChangeLog wording for key export
Remove return value from key export callback
Make key export callback and context connection-specific
Remove all occurrences of TLS < 1.2 PRF identifier
Remote key export identifier used for TLS < 1.2.
Add missing documentation for key export callback parameters
...
Where hashlen was previously ignored when the hash length could be
inferred from an md_alg parameter, the two must now match.
Adapt the existing tests accordingly. Adapt the sample programs accordingly.
This commit does not add any negative testing.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
The loop exits early iff there is a nonzero limb, so i==0 means that
all limbs are 0, whether the number of limbs is 0 or not.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Fix a bug introduced in "Fix multiplication producing a negative zero" that
caused the sign to be forced to +1 when A > 0, B < 0 and B's low-order limb
is 0.
Add a non-regression test. More generally, systematically test combinations
of leading zeros, trailing zeros and signs.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
In mbedtls_mpi_read_string, if the string is empty, return an empty bignum
rather than a bignum with one limb with the value 0.
Both representations are correct, so this is not, in principle, a
user-visible change. The change does leak however through
mbedtls_mpi_write_string in base 16 (but not in other bases), as it writes a
bignum with 0 limbs as "" but a bignum with the value 0 and at least one
limb as "00".
This change makes it possible to construct an empty bignum through
mbedtls_mpi_read_string, which is especially useful to construct test
cases (a common use of mbedtls_mpi_read_string, as most formats use in
production encode numbers in binary, to be read with mbedtls_mpi_read_binary
or mbedtls_mpi_read_binary_le).
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Fix mbedtls_mpi_mul_mpi() when one of the operands is zero and the
other is negative. The sign of the result must be 1, since some
library functions do not treat {-1, 0, NULL} or {-1, n, {0}} as
representing the value 0.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Shifting TA and TB before the loop is not necessary. If A != 0, it will be
done at the start of the loop iteration. If A == 0, then lz==0 and G is
correctly set to B after 0 loop iterations.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Fix a null pointer dereference in mbedtls_mpi_exp_mod(X, A, N, E, _RR) when
A is the value 0 represented with 0 limbs.
Make the code a little more robust against similar bugs.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
* development: (236 commits)
Changing the key length to 32 bytes in one of the PSA cipher setup tests
Removal of RC4 certs and fixes to docs and tests
Fix fd range for select on Windows
Refactor file descriptor checks into a common function
Removing global variable and moving variant function comment block
Fix typo in doc'n of session resumption API
Code review fixes
Fix warning in some configurations
Fix cmake build of fuzz_privkey
Fix async support in ssl_server2
Improve ChangeLog and migration guide entries
Use a proper DRBG in programs
Use the dedicated dummy_random in fuzzing programs
Fix cmake build of programs
Add ChangeLog and migration guide entries
Simplify internal code
Remove "internal RNG" code from ECP
Remove config option MBEDTLS_ECP_NO_INTERNAL_RNG
Add RNG params to private key parsing
Add RNG parameter to check_pair functions
...
Remove a case that cannot be triggered as PSA_ALG_SIGN_GET_HASH always
returns 0 for raw algorithms.
Signed-off-by: Janos Follath <janos.follath@arm.com>
PSA Crypto always passed MBEDTLS_MD_NONE to Mbed TLS, which worked well
as Mbed TLS does not use this parameter for anything beyond determining
the input lengths.
Some alternative implementations however check the consistency of the
algorithm used for pre-hash and for other uses in verification (verify
operation and mask generation) and fail if they don't match. This makes
all such verifications fail.
Furthermore, the PSA Crypto API mandates that the pre-hash and internal
uses are aligned as well.
Fixes#3990.
Signed-off-by: Janos Follath <janos.follath@arm.com>
Introduce psa_mac_compute_internal with an
additional `is_sign` parameter compared to
the psa_mac_compute API. The intent is to
call psa_mac_compute_internal() from
psa_mac_verify() as well to compute the
message MAC.
Signed-off-by: Ronald Cron <ronald.cron@arm.com>
Fix mbedtls_net_poll() and mbedtls_net_recv_timeout() often failing with
MBEDTLS_ERR_NET_POLL_FAILED on Windows: they were testing that the file
descriptor is in range for fd_set, but on Windows socket descriptors are not
limited to a small range. Fixes#4465.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Re-organize psa_mac_setup() to prepare the move
to a dedicated function of the additional checks
on the algorithm and the key attributes done by
this function. We want to move those checks in
a dedicated function to be able to do them
without duplicating them in psa_mac_compute().
Signed-off-by: Ronald Cron <ronald.cron@arm.com>
Reverting some deleted tests and changing the deprecated algo
Deleting deprecated headers from /alt-dummy dir
Corrections to the comments
Removal of deleted functions from compat-2.x.h
Corrections to tests/data_files/Makefile
Signed-off-by: TRodziewicz <tomasz.rodziewicz@mobica.com>
For TLS, secp256k1 is deprecated by RFC 8422 §5.1.1. For X.509,
secp256k1 is not deprecated, but it isn't used in practice, especially
in the context of TLS where there isn't much point in having an X.509
certificate which most peers do not support. So remove it from the
default profile. We can add it back later if there is demand.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Rewrite mbedtls_debug_print_mpi to be simpler and smaller. Leverage
mbedtls_mpi_bitlen() instead of manually looking for the leading
zeros.
Fix#4608: the old code made an invalid memory dereference when
X->n==0 (freshly initialized bignum with the value 0).
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
TLS used to prefer larger curves, under the idea that a larger curve has a
higher security strength and is therefore harder to attack. However, brute
force attacks are not a practical concern, so this was not particularly
meaningful. If a curve is considered secure enough to be allowed, then we
might as well use it.
So order curves by resource usage. The exact definition of what this means
is purposefully left open. It may include criteria such as performance and
memory usage. Risk of side channels could be a factor as well, although it
didn't affect the current choice.
The current list happens to exactly correspond to the numbers reported by
one run of the benchmark program for "full handshake/s" on my machine.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Upgrade the default list of hashes and curves allowed for TLS. The list is
now aligned with X.509 certificate verification: hashes and curves with at
least 255 bits (Curve25519 included), and RSA 2048 and above.
Remove MBEDTLS_TLS_DEFAULT_ALLOW_SHA1_IN_KEY_EXCHANGE which would no
longer do anything.
Document more precisely what is allowed by default.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Upgrade the default X.509 certificate verification profile
mbedtls_x509_crt_profile_default to the former value of
mbedtls_x509_crt_profile_next, which is hashes and curves with at least 255
bits (Curve25519 included), and RSA 2048 and above.
Document more precisely what goes into the default profile.
Keep the "next" profile unchanged for now.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
This makes it easier to ensure that crypto_spe.h is included everywhere it
needs to be, and that it's included early enough to do its job (it must be
included before any mention of psa_xxx() functions with external linkage,
because it defines macros to rename these functions).
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
We know that Montgomery multiplication will never be called without an
RNG, so make that clear from the beginning of the function.
Signed-off-by: Manuel Pégourié-Gonnard <manuel.pegourie-gonnard@arm.com>
This is necessary for the case where the public part of an EC keypair
needs to be computed from the private part - either because it was not
included (it's an optional component) or because it was compressed (a
format we can't parse).
This changes the API of two public functions: mbedtls_pk_parse_key() and
mbedtls_pk_parse_keyfile().
Tests and programs have been adapted. Some programs use a non-secure RNG
(from the test library) just to get things to compile and run; in a
future commit this should be improved in order to demonstrate best
practice.
Signed-off-by: Manuel Pégourié-Gonnard <manuel.pegourie-gonnard@arm.com>
- mbedtls_ecp_check_pub_priv() because it calls ecp_mul()
- mbedtls_pk_check_pair() because it calls the former
Signed-off-by: Manuel Pégourié-Gonnard <manuel.pegourie-gonnard@arm.com>
It was indirectly calling ecp_mul() without an RNG. That's actually the
rare case where this should be allowed, as ecp_muladd() is typically
used on non-secret data (to verify signatures or ZKPs) and documented as
not being constant-time.
Refactor a bit in order to keep the ability to call ecp_mul() without a
RNG, but not exposed publicly (except though muladd).
Signed-off-by: Manuel Pégourié-Gonnard <manuel.pegourie-gonnard@arm.com>
Fix trivial faulty calls in ECP test suite and ECP/ECJPAKE self-tests (by
adding a dummy RNG).
Several tests suites are not passing yet, as a couple of library
function do call ecp_mul() with a NULL RNG. The complexity of the fixes
range from "simple refactoring" to "requires API changes", so these will
be addressed in separate commits.
This makes the option MBEDTLS_ECP_NO_INTERNAL_RNG, as well as the whole
"internal RNG" code, obsolete. This will be addressed in a future
commit, after getting the test suites to pass again.
Signed-off-by: Manuel Pégourié-Gonnard <manuel.pegourie-gonnard@arm.com>
(This commit is best reviewed using `git show -b` as indentation levels
have changed.)
The documentation already states that the RNG parameter is mandatory,
since PRs #4488 and #4515. There are several families of functions to
consider here:
- private-key operations (sign, decrypt) all call
mbedtls_rsa_private() where this commit adds a non-NULL check;
- encrypt operations need an RNG for masking/padding and already had a
non-NULL check since #4515 (conditional on \p mode before that)
- verify operations no longer take an RNG parameter since #4515
So, after this commit, all RSA functions that accept an RNG will reach a
non-NULL check before the RNG is used.
Signed-off-by: Manuel Pégourié-Gonnard <manuel.pegourie-gonnard@arm.com>
Removes conditional code blocks relating
to MBEDTLS_SSL_TRUNCATED_HMAC from ssl_cli.c
and ssl_srv.c.
Signed-off-by: Thomas Daubney <thomas.daubney@arm.com>
Removes conditional code compilation blocks
and code paths relating to the
MBEDTLS_SSL_TRUNCATED_HMAC config option.
Signed-off-by: Thomas Daubney <thomas.daubney@arm.com>
Implement one-shot MAC APIs, psa_mac_compute and psa_mac_verify, introduced in PSA Crypto API 1.0.
Signed-off-by: gabor-mezei-arm <gabor.mezei@arm.com>
mbedtls_dhm_get_value can be seen as either a copy function or a getter
function. Given the name and the semantics, it's more of a getter, even if
it "gets" by doing a copy. Therefore, put the context first, and the
selector next, leaving the output for last.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Add two functions mbedtls_dhm_get_len() and mbedtls_dhm_get_bitlen() to
query the length of the modulus in bytes or bits.
Remove the len field: the cost of calling mbedtls_dhm_get_len() each time
it's needed is negligible, and this improves the abstraction of the DHM
module.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
When MBEDTLS_RSA_ALT is defined, mbedtls_rsa_context does not necessarily
have fields called N and E of type mbedtls_mpi. Don't emit pk_debug
information with MBEDTLS_RSA_ALT. This is not ideal but at least the
library compiles.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
library/ecp_alt.h (declaring individual functions of the ECP module that can
be substituted, included when building the library with
MBEDTLS_ECP_INTERNAL_ALT enabled) clashes with ecp_alt.h (not provided,
declaring types of the ECP module when the whole implementation is
substituted, included when building the library with MBEDTLS_ECP_ALT enabled).
Depending on the search path during build, this can make MBEDTLS_ECP_ALT
unusable.
Rename library/ecp_alt.h to follow the naming convention of other alt headers:
MBEDTLS_XXX_ALT corresponds to xxx_alt.h.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Conflicts:
include/mbedtls/ssl.h
include/psa/crypto_struct.h
Conflicts fixed by using the code from development branch
and manually re-applying the MBEDTLS_PRIVATE wrapping.
Unrelated to RSA (only used in ECP), but while improving one
mbedtls_safe_cond_xxx function, let's improve the other as well.
Signed-off-by: Manuel Pégourié-Gonnard <manuel.pegourie-gonnard@arm.com>
mbedtls_mpi_cf_bool_eq() is a verbatim copy of mbedtls_ssl_cf_bool_eq()
Deduplication will be part of a future task.
Signed-off-by: Manuel Pégourié-Gonnard <manuel.pegourie-gonnard@arm.com>
Remove padding parameters as mbedtls_rsa_init()
cannot return an error code when padding
parameters are invalid.
Signed-off-by: Ronald Cron <ronald.cron@arm.com>
mbedtls_rsa_set_padding() now returns the error
code MBEDTLS_ERR_RSA_INVALID_PADDING when
padding parameters are invalid.
Signed-off-by: Ronald Cron <ronald.cron@arm.com>
This change enables automatic detection and consumption of Mbed TLS
library targets from within other CMake projects. By generating an
`MbedTLSConfig.cmake` file, consuming projects receive a more complete
view of these targets, allowing them to be used as dependencies which
properly inherit the transitive dependencies of the libraries.
This is fairly fragile, as it seems Mbed TLS's libraries do not appear
to properly model their dependencies on other targets, including
third-party dependencies. It is, however, sufficient for building and
linking the compiled Mbed TLS libraries when there are no third-party
dependencies involved. Further work is needed for more complex
use-cases, but this will likely meet the needs of most projects.
Resolves#298. Probably useful for #2857.
Signed-off-by: Chris Kay <chris.kay@arm.com>
Calling mbedtls_mpi_cmp_int reveals the number of leading zero limbs
to an adversary who is capable of very fine-grained timing
measurements. This is very little information, but could be practical
with secp521r1 (1/512 chance of the leading limb being 0) if the
adversary can measure the precise timing of a large number of
signature operations.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
ssl_server2 had a check that we never try to use a minor version lower
than 2 with DTLS, but that check is no longer needed, as there's no way
that would happen now that MBEDTLS_SSL_MINOR_VERSION_1 is no longer
public.
Signed-off-by: Manuel Pégourié-Gonnard <manuel.pegourie-gonnard@arm.com>
This commit removes the API
```
mbedtls_ssl_conf_ciphersuites_for_version()
```
which allows to configure lists of acceptable ciphersuites
for each supported version of SSL/TLS: SSL3, TLS 1.{0,1,2}.
With Mbed TLS 3.0, support for SSL3, TLS 1.0 and TLS 1.1
is dropped. Moreover, upcoming TLS 1.3 support has a different
notion of cipher suite and will require a different API.
This means that it's only for TLS 1.2 that we require
a ciphersuite configuration API, and
```
mbedtls_ssl_conf_ciphersuites()
```
can be used for that. The version-specific ciphersuite
configuration API `mbedtls_ssl_conf_ciphersuites_for_version()`,
in turn, is no longer needed.
Signed-off-by: Hanno Becker <hanno.becker@arm.com>
Signed-off-by: Manuel Pégourié-Gonnard <manuel.pegourie-gonnard@arm.com>
The idiom "resize an mpi to a given size" appeared 4 times. Unify it
in a single function. Guarantee that the value is set to 0, which is
required by some of the callers and not a significant expense where
not required.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Since the internal function mpi_fill_random_internal() assumes that X
has the right size, there is no need to call grow().
To further simplify the function, set the sign outside, and zero out
the non-randomized part directly.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
In real life, min << N and the probability that mbedtls_mpi_random()
fails to find a suitable value after 30 iterations is less than one in
a billion. But at least for testing purposes, it's useful to not
outright reject "silly" small values of N, and for such values, 30
iterations is not enough to have a good probability of success.
Pick 250 iterations, which is enough for cases like (min=3, N=4), but
not for cases like (min=255, N=256).
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
This comment is no longer in the specific context of generating a
random point on an elliptic curve.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
mbedtls_mpi_random() uses mbedtls_mpi_cmp_mpi_ct(), which requires its
two arguments to have the same storage size. This was not the case
when the upper bound passed to mbedtls_mpi_random() had leading zero
limbs.
Fix this by forcing the result MPI to the desired size. Since this is
not what mbedtls_mpi_fill_random() does, don't call it from
mbedtls_mpi_random(), but instead call a new auxiliary function.
Add tests to cover this and other conditions with varying sizes for
the two arguments.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Instead of generating blinding values and keys in a not-quite-uniform way
(https://github.com/ARMmbed/mbedtls/issues/4245) with copy-pasted code,
use mbedtls_mpi_random().
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
dhm_make_common includes a piece of code that is identical to
dhm_random_below except for returning a different error code in one
case. Call dhm_random_below instead of repeating the code.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
P-1 is as bad as 1 as a blinding value. Don't accept it.
The chance that P-1 would be randomly generated is infinitesimal, so
this is not a practical issue, but it makes the code cleaner. It was
inconsistent to accept P-1 as a blinding value but not as a private key.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Unify the common parts of mbedtls_dhm_make_params and mbedtls_dhm_make_public.
No intended behavior change, except that the exact error code may
change in some corner cases which are too exotic for the existing unit
tests.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Instead of generating blinding values in a not-quite-uniform way
(https://github.com/ARMmbed/mbedtls/issues/4245) with copy-pasted
code, use mbedtls_mpi_random().
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Since mbedtls_mpi_random() is not specific to ECC code, move it from
the ECP module to the bignum module.
This increases the code size in builds without short Weierstrass
curves (including builds without ECC at all) that do not optimize out
unused functions.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Rename mbedtls_ecp_gen_privkey_sw to mbedtls_mpi_random since it has
no particular connection to elliptic curves beyond the fact that its
operation is defined by the deterministic ECDSA specification. This is
a generic function that generates a random MPI between 1 inclusive and
N exclusive.
Slightly generalize the function to accept a different lower bound,
which adds a negligible amount of complexity.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
mbedtls_ecp_gen_privkey_mx generates a random number with a certain
top bit set. Depending on the size, it would either generate a number
with that top bit being random, then forcibly set the top bit to
1 (when high_bit is not a multiple of 8); or generate a number with
that top bit being 0, then set the top bit to 1 (when high_bit is a
multiple of 8). Change it to always generate the top bit randomly
first.
This doesn't make any difference in practice: the probability
distribution is the same either way, and no supported or plausible
curve has a size of the form 8n+1 anyway. But it slightly simplifies
reasoning about the behavior of this function.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Don't calculate the bit-size of the initially generated random number.
This is not necessary to reach the desired distribution of private
keys, and creates a (tiny) side channel opportunity.
This changes the way the result is derived from the random number, but
does not affect the resulting distribution.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
The library rejected an RNG input of all-bits-zero, which led to the
key 2^{254} (for Curve25519) having a 31/32 chance of being generated
compared to other keys. This had no practical impact because the
probability of non-compliance was 2^{-256}, but needlessly
complicated the code.
The exception was added in 98e28a74e3 to
avoid the case where b - 1 wraps because b is 0. Instead, change the
comparison code to avoid calculating b - 1.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
For Montgomery keys, n_bits is actually the position of the highest
bit and not the number of bits, which would be 1 more (fence vs
posts). Rename the variable accordingly to lessen the confusion.
No semantic change.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
