Fix a race condition in `mbedtls_aes_ni_has_support()` with some compilers.
A compiler could hoist the assignment `done = 1` above the assignment to `c`,
in which case if two threads call `mbedtls_aes_ni_has_support()` at almost
the same time, they could be interleaved as follows:
Initially: done = 0, c = 0
thread A thread B
if (!done)
done = 1; # hoisted
if (!done)
return c & what; # wrong!
c = cpuid();
return c & what
This would lead to thread B using software AES even though AESNI was
available. This is a very minor performance bug. But also, given a very
powerful adversary who can block thread A indefinitely (which may be
possible when attacking an SGX enclave), thread B could use software AES for
a long time, opening the way to a timing side channel attack.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
This is mostly unrelated to other commits in this PR, except for the
fact that one of the added X.509 tests revealed that with UBSan.
Signed-off-by: Manuel Pégourié-Gonnard <manuel.pegourie-gonnard@arm.com>
That way when API are declared, the types they use are defined.
This should resolve the issues related to psa_xyz_init functions
returning a structure described in #7087.
Signed-off-by: Ronald Cron <ronald.cron@arm.com>
When passed a zero-length val, the function was free-ing the buffer as
the documentation suggests:
* \param val_len The minimum length of the data buffer needed.
* If this is 0, do not allocate a buffer for the associated
* data.
* If the OID was already present, enlarge, shrink or free
* the existing buffer to fit \p val_len.
However it kept the previous length, leaving the val structure in the
corresponding item in the output list in an inconsistent state:
p == NULL but len != 0
As a result, functions that would try using this item in the list
(including the same function!) afterwards would trip an dereference the
NULL pointer.
Signed-off-by: Manuel Pégourié-Gonnard <manuel.pegourie-gonnard@arm.com>
The commented out tests cause crashes (in different ways) until the bug
is fixed; the first two test are passing already and are here mostly to
provide a reference point.
The bug report was using programs/x509/cert_write, but string_to_names()
is what it was really targetting, which is better for automated tests.
The strings used are a minor adapation of those from the report.
Signed-off-by: Manuel Pégourié-Gonnard <manuel.pegourie-gonnard@arm.com>
I'm just trying to have a shorter name to avoid repeating a long
expression. This is a job for a pointer, not copying a struct.
Signed-off-by: Manuel Pégourié-Gonnard <manuel.pegourie-gonnard@arm.com>
This version is incomplete. I failed to noticed it when adding a more
complete version, making the existing one redundant.
Signed-off-by: Manuel Pégourié-Gonnard <manuel.pegourie-gonnard@arm.com>
Since Mbed TLS 3.0, blinding is no longer optional in ECDSA.
`mbedtls_ecdsa_write_signature()` and
`mbedtls_ecdsa_write_signature_restartable()` error out if
`f_rng == NULL`. We forgot to update the function documentation.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
In public functions that take `f_rng, p_rng` callbacks, link to the
documentation of the callback which is attached to the type name
`mbedtls_f_rng_t`.
Resolves#5868.
```
grep -l -w 'f_rng)' include | xargs perl -i -pe 's/\Qint (*f_rng)(void *, unsigned char *, size_t)\E/mbedtls_f_rng_t *f_rng/g'
```
and include `platform_util.h` where needed.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
