We do not intend to support multithreaded testing in 2.28, so
introducing a C11 feature here is an unnecessary burden.
Signed-off-by: David Horstmann <david.horstmann@arm.com>
Make sure that extra UNPOISON calls do not cause the poisoning counter
to underflow and wrap around.
Memory that is unpoisoned multiple times should remain unpoisoned.
Signed-off-by: David Horstmann <david.horstmann@arm.com>
Some platforms may support ASan but be C99-only (no C11 support).
These platforms will support ASan metatests but not memory poisoning,
which requires C11 features.
To allow for this, create a separate platform requirement, "poison",
in metatest.c to distinguish generic ASan metatests from ones that
require suppport for memory poisoning.
In practice our platforms support both, so run "poison" tests in
the same all.sh components where we run "asan" ones.
Signed-off-by: David Horstmann <david.horstmann@arm.com>
This allows unusually-nested memory poisoning to work correctly, since
it keeps track of whether any buffers are still poisoned, rather than
just disabling poisoning at the first call to the UNPOISON() macro.
Signed-off-by: David Horstmann <david.horstmann@arm.com>
Allow memory poisoning to be enabled and disabled at runtime using a
thread-local flag. This allows poisoning to be disabled whenever a PSA
function is called but not through the test wrappers, removing false
positive use-after-poisons.
Signed-off-by: David Horstmann <david.horstmann@arm.com>
`psa_collect_statuses.py` runs `make RECORD_PSA_STATUS_COVERAGE_LOG=1`,
which builds with `RECORD_PSA_STATUS_COVERAGE_LOG`. In this mode, the build
includes wrappers for PSA functions, which conflict with the newly
introduced wrappers that are enabled whenever `MBEDTLS_TEST_HOOKS` is
enabled. In the future, the collect-statuses mechanism should use the new
generic wrapper mechanism. For the time being, keep the old wrappers and
avoid the new wrappers when doing the collect-statuses build.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
For now, only instrument the one function for which buffer copying has been
implemented, namely `psa_cipher_encrypt`.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Code in unit tests (`tests/suites/*.function`) and in test support
code (`tests/src/**.c`) will now go through the wrapper functions when they
call a PSA API function and `MBEDTLS_TEST_HOOKS` is enabled.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Commit files generated by `tests/scripts/generate_psa_wrappers.py`. As of
this commit, the new code is neither useful (the wrappers just call the
underlying functions) nor used (the wrapper functions are not called from
anywhere). This will change in subsequent commits.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
The new script `tests/scripts/generate_psa_wrappers.py` generates the
implementation of wrapper functions for PSA API functions, as well as a
header that defines macros that redirect calls to the wrapper functions. By
default, the wrapper functions just call the underlying library function.
With `--log`, the wrapper functions log the arguments and return values.
This commit only introduces the new script. Subsequent commits will
integrate the wrappers in the build.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Make sure that we don't enable memory poisoning when
MBEDTLS_PSA_COPY_CALLER_BUFFERS is disabled.
Signed-off-by: David Horstmann <david.horstmann@arm.com>
This check is intended to ensure that we do not write intermediate
results to the shared output buffer. This check will be made obselete
by generic memory-poisoning-based testing for all functions.
Signed-off-by: David Horstmann <david.horstmann@arm.com>
Since we are applying hooks transparently to all tests, we cannot setup
and teardown test hooks in the tests. Instead we must do this in the
test wrappers which are used to pre-poison and unpoison memory.
Signed-off-by: David Horstmann <david.horstmann@arm.com>
Use the preprocessor to wrap psa_cipher_encrypt in a wrapper that
poisons the input and output buffers.
Signed-off-by: David Horstmann <david.horstmann@arm.com>
These are needed to allow them to operate on buffer copies without
triggering ASan use-after-poison detection.
Signed-off-by: David Horstmann <david.horstmann@arm.com>
This test case exercises an integer overflow in calloc. Under Asan, with
a modern Clang, this triggers an Asan complaint. The complaint can be
avoided with ASAN_OPTIONS=allocator_may_return_null=1, but this has to
be set in the environment before the program starts, and could hide
other errors.
Reproduced from a commit in mbedtls#8286
Signed-off-by: David Horstmann <david.horstmann@arm.com>
Asan poisons memory with an 8-byte granularity. We want to make sure that
the whole specified region is poisoned (our typical use case is a
heap-allocated object, and we want to poison the whole object, and we don't
care about the bytes after the end of the object and up to the beginning of
the next object). So align the start and end of the region to (un)poison to
an 8-byte boundary.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
While an area of memory is poisoned, reading or writing from it triggers a
sanitizer violation.
Implemented for ASan.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Same test cases as in the previous commit.
Remove the redundant symmetric crypto dependency.
The dependency is ensured by the fact that:
1) the test case forces a cipher suite
2) ssl-opt.sh enforces automatically that the
forced ciphersuite is available.
3) The fact that the forced ciphersuite is
available implies that the symmetric
cipher algorithm it uses is available as
well.
Signed-off-by: Ronald Cron <ronald.cron@arm.com>
Fix some dependencies on symmetric crypto that
were not correct in case of driver but not
builtin support. Revealed by "Analyze driver
test_psa_crypto_config_accel_cipher_aead vs reference
test_psa_crypto_config_reference_cipher_aead" in
analyze_outcomes.py.
Signed-off-by: Ronald Cron <ronald.cron@arm.com>
