lcov had a bug whereby it tries to create the output file relative to /
if it has emitted a warning. We do CI runs on Ubuntu 16.04 which is too
old to have the fix. As a quick fix for the CI, work around the bug.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
"Key generation method" was misleading since it also applies to key
derivation. Change "key generation" to "key production", which we aren't
using yet and has roughly the right intuition. Change "method" to
"parameters" which there seems to be a slight preference for. Discussion
thread: https://github.com/Mbed-TLS/mbedtls/pull/8815#discussion_r1486524295
Identifiers renamed:
psa_key_generation_method_t → psa_key_production_parameters_t
psa_key_generation_method_s → psa_key_production_parameters_s
PSA_KEY_GENERATION_METHOD_INIT → PSA_KEY_PRODUCTION_PARAMETERS_INIT
method → params
method_data_length → params_data_length
default_method → default_production_parameters
psa_key_generation_method_is_default → psa_key_production_parameters_are_default
setup_key_generation_method → setup_key_production_parameters
key_generation_method_init → key_production_parameters_init
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
In the driver wrapper for psa_generate_key() and psa_generate_key_ext():
* Invoke the built-in code if using a non-default method, even if there
might be an accelerator. This is ok because we only support non-default
methods for RSA and we don't support driver-only RSA, therefore a
non-default method will always have built-in code behind it.
* Return NOT_SUPPORTED if trying to use a non-default method with an opaque
driver.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Instead of passing the size of the whole structure, just pass the data
length and let the implementation worry about adding the size of the
structure. The intent with passing the structure size was to allow
the client code in a client-server implementation to know nothing
about the structure and just copy the bytes to the server. But that was not
really a useful consideration since the application has to know the
structure layout, so it has to be available in the client implementation's
headers. Passing the method data length makes life simpler for everyone by
not having to worry about possible padding at the end of the structure, and
removes a potential error condition
(method_length < sizeof(psa_key_generation_method_t)).
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Fix // comments stopping on 'n' instead of newlines. Also allow
backslash-newline in // comments.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
The Base class generates trivial wrappers that just call the underlying
function. It is meant as a base class to construct useful wrapper generators.
The Logging class generates wrappers that can log the inputs and outputs to
a function.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
secp224k1 is the one with 225-bit private keys.
The consequences of this mistake were:
* We emitted positive test cases for hypothetical SECP_R1_225 and
SECP_K1_224 curves, which were never executed.
* We emitted useless not-supported test cases for SECP_R1_225 and SECP_K1_224.
* We were missing positive test cases for SECP_R1_224 in automatically
generated tests.
* We were missing not-supported test cases for SECP_R1_224 and SECP_K1_225.
Thus this didn't cause test failures, but it caused missing test coverage
and some never-executed test cases.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
`$(MBEDTLS_TEST_OBJS)` included TLS-specific test support modules in
`tests/Makefile` but not in `programs/Makefile`. This difference is not
actually necessary. What is necessary is that all programs that use
functions from TLS-specific test support modules are linked with those
modules in addition to `-lmbedtls`, and programs that are not linked with
`-lmbedtls` are not linked with TLS-specific test support modules. Since we
always pass `-lmbedtls` when linking programs in `programs/Makefile`, we can
link with the TLS-specific test support modules as well. This keeps things
simpler.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
fixup "Create common.make with LOCAL_CFLAGS and friends"
The code wasn't what I had intended, although it was functionally
equivalent. Make it more readable and more robust.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
When building with Make on a Unix-like platform (shell and compiler),
auto-detect configurations that may require linking with pthread.
This removes the need for MAKE_THREADING_FLAGS in all.sh.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Create a common.make for definitions that are shared between tests/Makefile
and programs/Makefile, to facilitate maintenance. Start populating it with
CFLAGS/LDFLAGS variables. More to follow in subsequent commits.
Keep library/Makefile independent, at least for the time being.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Add new function to build_tree.py to return the crypto
name for the project; either tfpsacrypto or mbedcrypto.
Deploy this function where needed.
Signed-off-by: Thomas Daubney <thomas.daubney@arm.com>
Add crypto_core_directory in build_tree.py so that
the libary/core directory can be returned based
on what repository we are in.
Signed-off-by: Thomas Daubney <thomas.daubney@arm.com>
Rename some variables in generate_driver_wrappers.py
now that the script has to work in two repositories
as opposed to just mbed tls.
Signed-off-by: Thomas Daubney <thomas.daubney@arm.com>
Compiling a regex improves performance and avoids accidentally
combining it with a string. This commit makes this change.
Signed-off-by: Yanray Wang <yanray.wang@arm.com>
lcov.sh can now be called from any build directory and
also still works with in-place builds too.
Signed-off-by: Thomas Daubney <thomas.daubney@arm.com>
Add repository detection (credit to davidhorstmann-arm
for adding this in all.sh previously) and use repository
detection to set the library directory and title
variables.
Signed-off-by: Thomas Daubney <thomas.daubney@arm.com>
All paths in both switch statement lead to a return, therefore the
end block in both of these functions can never be reached. Fix this by
making sure the end block is always called (set status rather than just
return), as its safer for future changes. Found by coverity scan.
Signed-off-by: Paul Elliott <paul.elliott@arm.com>
Move get_key_buf_size/get_builtin_key out of
the psa wrapper auto generated file
Slot_management.c include the head file instead of the source file
Signed-off-by: Xiaokang Qian <xiaokang.qian@arm.com>
The shipped Visual Studio project files were misconfigured for build
combinations of 64 bit Release builds.
Signed-off-by: Simon Butcher <simon.butcher@arm.com>
CryptGenRandom and lstrlenW are not permitted in Windows Store apps,
meaning apps that use mbedTLS can't ship in the Windows Store.
Instead, use BCryptGenRandom and wcslen, respectively, which are
permitted.
Also make sure conversions between size_t, ULONG, and int are
always done safely; on a 64-bit platform, these types are different
sizes.
Also suppress macro redefinition warning for intsafe.h:
Visual Studio 2010 and earlier generates C4005 when including both
<intsafe.h> and <stdint.h> because a number of <TYPE>_MAX constants
are redefined. This is fixed in later versions of Visual Studio.
The constants are guaranteed to be the same between both files,
however, so we can safely suppress the warning when including
intsafe.h.
Signed-off-by: Kevin Kane <kkane@microsoft.com>
Now that p256-m is officially a production feature and not just an example,
give it a more suitable name.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
The official spelling of the trade mark changed from all-lowercase "mbed"
to normal proper noun capitalization "Mbed" a few years ago. We've been
using the new spelling in new text but still have the old spelling in a
lot of text. This commit updates most occurrences of "mbed TLS":
```
sed -i -e 's/mbed TLS/Mbed TLS/g' $(git ls-files ':!ChangeLog' ':!tests/data_files/**' ':!tests/suites/*.data' ':!programs/x509/*' ':!configs/tfm*')
```
Justification for the omissions:
* `ChangeLog`: historical text.
* `test/data_files/**`, `tests/suites/*.data`, `programs/x509/*`: many
occurrences are significant names in certificates and such. Changing
the spelling would invalidate many signatures and tests.
* `configs/tfm*`: this is an imported file. We'll follow the upstream
updates.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Previously the code only recognized the old spelling "mbed TLS", so it
missed doxygen/input/doc_mainpage.h.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
In case of an out-of-tree build with the CMake build
system the path to crypto_config.h has to be defined
as the path to mbedtls_config.h. Add this possibility
tp generate_query_config.pl.
Signed-off-by: Ronald Cron <ronald.cron@arm.com>
Boolean options that modify the behavior of a module are supposed to be in
the "feature support" section, not in the "configuration options" support:
that section is documented to contain commented-out definitions with a
value, for which the comment contains the default version. In particular,
merely uncommenting a definition in the "configuration options" section is
not supposed to change anything.
Move the offending boolean options to the proper section.
This causes those options to be enabled by `config.py full` unless
explicitly excluded. For all the offending options, this is undesirable, so
make sure those options are indeed excluded.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Conflicts:
* `include/psa/crypto_sizes.h`: the addition of the `u` suffix in this branch
conflicts with the rework of the calculation of `PSA_HASH_MAX_SIZE` and
`PSA_HMAC_MAX_HASH_BLOCK_SIZE` in `development`. Use the new definitions
from `development`, and add the `u` suffix to the relevant constants.
Check for the 'drivers' and 'programs' directories additionally to the
ones that are already there.
Signed-off-by: David Horstmann <david.horstmann@arm.com>
They aren't used anywhere.
Keep the command line options of all.sh to avoid breaking any wrapper
scripts that people might have.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
This reverts commit eb2c39ed2bc6a126ae7bdb6eab0457ebd6a32cfc.
We temporarily turned off the cryptography requirement on Windows due to a
CI instance that had an old, incompatible verison of pip. That CI instance
has been upgraded so we no longer need the workaround.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Some basic test coverage for now:
* Nominal operation.
* Larger output buffer.
* Clone an operation and use it after the original operation stops.
Generate test data automatically. For the time being, only do that for
hashes that Python supports natively. Supporting all algorithms is future
work.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
'Removed' and 'NotCreated' should be displayed in new and old column
respectively. The value of delta is reflected on change column. This
commit handles the corner cases properly.
Signed-off-by: Yanray Wang <yanray.wang@arm.com>
To keep a concise markdown table, we don't list text and data
section from old Git revision. However, it should be ideal to keep
those two sections in CSV output. Therefore, we list comparison
result for CSV output in following format:
filename new(text) new(data) old(text) old(data) change(text) change(data)
Additionally, if a file only exits in new Git revision not in old
Git revision, it's marked as `NotCreated` as we haven't created this
file yet from perspective of old Git revision.
Signed-off-by: Yanray Wang <yanray.wang@arm.com>
In comparison result, to indicate it's a delta value, we add `+` in
front of positive values. For unchanged attributes, it's still
shown as `0'.
Signed-off-by: Yanray Wang <yanray.wang@arm.com>
Previously we set loglevel >= WARNING printed to stderr and
loglevel < WARNING printed to stdout. To be more flexible, we
replace this `WARNING` value with an argument: split_level and leave
`WARNING` as default split_level if not set.
Signed-off-by: Yanray Wang <yanray.wang@arm.com>
Transparent drivers don't have this entry points. Other entry points
that are only for opaque drivers are not listed here but only in the
opaque_driver.json file, so this was likely a mistake.
Signed-off-by: Manuel Pégourié-Gonnard <manuel.pegourie-gonnard@arm.com>
The EXCLUDE_FROM_FULL list in config.py should be, and used to be, but
no longer is, in alphabetical order, and with a comma on the last
element.
Signed-off-by: Janos Follath <janos.follath@arm.com>
By default, the full configuration enables all options. But we
specifically don't want to enable MBEDTLS_ECP_WITH_MPI_UINT except where
it's done explicitly, because it disables the old ecp. So it needs to be
added to the exceptions in config.py (EXCLUDE_FROM_FULL).
Signed-off-by: Janos Follath <janos.follath@arm.com>
The latest and earliest clang/GCC compilers are now used through
variables instead of symlinks and also the all.sh script is updated
to support options for overriding the default values.
Signed-off-by: Gowtham Suresh Kumar <gowtham.sureshkumar@arm.com>
When '--show-all' is enabled, all the objects will be displayed in
comparison result no matter if there is code size change or not.
Signed-off-by: Yanray Wang <yanray.wang@arm.com>
It makes sense to display code size changes if a file has been
removed in our library. With this commit we track old objects as
well. If a file is not present in the new Git revision,
we display -old_size in the new_size column. The size change is
marked as `Removed` to indicate the file has been removed.
Signed-off-by: Yanray Wang <yanray.wang@arm.com>
The result format for code size comparison is:
filename new(text) new(data) change(text) change(data)
yyy.o xxx xxx xx xx
The numbers followed are in bytes.
Signed-off-by: Yanray Wang <yanray.wang@arm.com>
Percentage is not a useful number when looking at code size changes.
Since it depends on the base of the code size. It might give
misleading information by simply looking at the numbers.
Signed-off-by: Yanray Wang <yanray.wang@arm.com>
The Ubuntu 16.04 and 22.04 docker images have been updated with
earliest and latest versions of gcc and clang respectively.
This patch adds the necessary component and support functions
required for the CI to run these compilers.
For FreeBSD we invoke the function by name so a condition is added
to disable the existing test_clang_opt function for linux.
Signed-off-by: Gowtham Suresh Kumar <gowtham.sureshkumar@arm.com>
