This property specifies the flag used to pass the linker script filename
through the compiler front end tot he linker.
For clang, we use the general purpose linker-pass through -Wl flag with -T:
-Wl,-T as clang doesn't support -T.
For gcc, we use -T directly as this keeps the picolibc specs file from
inserting the picolibc linker script as well.
If the compiler doesn't specify a value, we continue to use -Wl,-T as
before.
Signed-off-by: Keith Packard <keithp@keithp.com>
When building a native library there is a few options
we cannot pass to the compiler and linker,
including instructing them to genrate non PIE code
(as it is still to early to say that),
or garbage collect unused sections (as we are not yet doing
the final linking).
We also need to provide different
link options when building the elf for the DTS gen_handles
parsing (as that script requires a "final" executable elf)
than when we build the native library itself.
Signed-off-by: Alberto Escolar Piedras <alberto.escolar.piedras@nordicsemi.no>
So they depend or select on the right NATIVE_BUILD
instead of NATIVE_APPLICATION.
Signed-off-by: Alberto Escolar Piedras <alberto.escolar.piedras@nordicsemi.no>
Device dependencies are not always required, so make them optional via
CONFIG_DEVICE_DEPS. When enabled, the gen_device_deps script will run so
that dependencies are collected and part of the final image. Related
APIs will be also made available. Since device dependencies are used in
just a few places (power domains), disable the feature by default. When
not enabled, a second linking pass will not be required.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Rename the Kconfig option to be in line with recent renamings in device
handles/dependencies.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Use the "device_deps" naming scheme to emphasize we are storing device
dependencies. The fact we are using device handles to store them is an
implementation detail.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Rename the gen_handles script and all of its references/associated files
to gen_device_deps. The new new makes things more clear, because the
script just take care of generating, for each device, an array of device
dependencies. While device handles are used internally to store this
information, it is in reality an implementation detail.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
The syscall generation phase parses all header files to look
for potential syscalls, and emits all the relevant files to
enable syscalls. However, this results in all the syscall
marshalling functions being included in the final binary.
This is due to these functions being referred to inside
the dispatch list, resulting in ineffective garbage
collection during linking. Previous commits allows each
drivers and subsystems to specify which header files
containing syscalls are relevant. So this commit changes
the syscall generation to only include the syscalls needed
for the build in the syscall dispatch list and removing
various bits related to that. This allows the linker to
garbage collect unused syscall related function, and thus
reducing final binary size.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
When CONFIG_HAS_DYNAMIC_DEVICE_HANDLES is selected, devices handles are
placed in RAM region so that they can be modified at runtime. However,
the gen_handles.py script added the `const` attribute to the device
handle arrays regardless of this setting, leading to faults when running
the application. This may be an indicator that this feature is not
actively used.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
This variable is only needed when gen_handles.py is called, so define it
under the if() above.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Add an option for running the build time initialization priority check
as part of the build.
By default the build is going to fail if any error is reported
(dependent devices, inverted priority), optionally it can also be failed
on warnings (dependent devices, same priority).
Signed-off-by: Fabio Baltieri <fabiobaltieri@google.com>
The current approach of failing the build on ztest with no optimization
broke coverage builds, and generally raised some concerns about being
too aggressive.
Downgrade the error to a warning and rework the option to inhibit the
warning, while also dropping it automatically for POSIX (that are not
really affected by stack size) and coverage run (that always runs with
no optimization).
Will reconsider this down the road if we still see issues filed for the
tests broken with no optimization and no further tuning.
Signed-off-by: Fabio Baltieri <fabiobaltieri@google.com>
Align Zephyr modules with other user facing variables where settings
can be defined or extended, meaning Zephyr modules now supports:
ZEPHYR_MODULES and EXTRA_ZEPHYR_MODULES.
Support for ZEPHYR_EXTRA_MODULES is kept foir backward compatibility.
Signed-off-by: Torsten Rasmussen <Torsten.Rasmussen@nordicsemi.no>
Many tests are known to fail when built and no compiler optimizations.
Add a CMake check to error out when building a ztest based test with no
optimization, ask not file issues about it but also adds an opt-out
option to bypass the error for tests are actually designed to work in
this setup.
Signed-off-by: Fabio Baltieri <fabiobaltieri@google.com>
Allow applications to create a VERSION file in their application folder.
The presence of a VERSION file will create a app_version_h target which
generates an app_version.h file using same infrastructure as the Zephyr
kernel version.h generation.
Signed-off-by: Torsten Rasmussen <Torsten.Rasmussen@nordicsemi.no>
Generalize the VERSION and version.h generation so that the same
infrastructure can be reused for generating other version related
header files, such as an application version header.
Signed-off-by: Torsten Rasmussen <Torsten.Rasmussen@nordicsemi.no>
Add a new compiler option to control tracking locations of tokens across
macro expansions. It may be useful to disable it when debugging long
macro expansion chains.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Seems it may have been lost during refactoring
that C++ code should also be built as not
position independent code to save space.
Let's enable it just like for plain C.
Signed-off-by: Alberto Escolar Piedras <alberto.escolar.piedras@nordicsemi.no>
This function allows subsystems to define libraries which get added to the
link command after all other libraries and modules. It's useful when using
a toolchain library, like libc or libgcc, as those can get added when
processing the 'lib' directory, before any module libraries and hence might
not get used to resolve symbols from modules.
Signed-off-by: Keith Packard <keithp@keithp.com>
Add an explicit compiler configuration, COMPILER_FREESTANDING, which
controls whether the compiler should operate in freestanding or hosted mode
(according to the C and C++ language specifications.
This depends on having a C library which conforms with the language
specification, and the minimal C library does not. Have the minimal C
library select COMPILER_FREESTANDING to continue using freestanding mode
with that library.
For other C libraries, leave this disabled by default while allowing users
to enable it if they want to go back to the previous configuration.
Signed-off-by: Keith Packard <keithp@keithp.com>
Add CONFIG_BUILD_OUTPUT_STRIP_PATHS to allow keeping absolute paths
in the build output.
This can be helpful for IDEs to detect links in a console output and
jump to the code location (e.g. when running tests in native_posix
locally).
Debuggers usually have a path mapping feature to ensure the files are
still found.
Signed-off-by: Martin Jäger <martin@libre.solar>
Add a Kconfig option to set the compiler save-temps flag and set the GCC
implementation. This is very useful for troubleshooting macro expansion
issues, having an option allows a user to set it like any other config
option.
Signed-off-by: Fabio Baltieri <fabiobaltieri@google.com>
This adds a new linker property specifically for passing
"-no-pie" to linker. Older binutils' LD (<= 2.36) do not
support this flag and will behave erratically if set. It
would parse "-no-pie" separately as "-n" and "-o-pie",
which would result in the output file being "-pie"
instead of "zephyr*.elf". Moreover, LLVM lld does not
support -no-pie but --no-pie (note the extra hyphen).
By having no-pie as a linker property, we can pass
correct no-pie flag to these linkers (or none at all).
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
Allows extended or replacing the default MCUboot signing
functionality by using a cmake property which can be set by modules
if alternate or modified signing functionality is required.
Signed-off-by: Jamie McCrae <jamie.mccrae@nordicsemi.no>
There is no need to generate all the *_mrsh.c files for
marshalling syscall arguments when userspace is not enabled.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
The oneApi support has bit rotten since it was first introduced. Update
the support to function with the latest 2023.0.0 release and add a
check to only support that version or newer for now. Versions before
2021.2.0 have linker script failures.
Various fixes made:
* In the 2023.0.0 release, various binaries are in a llvm-bin path so
add support to search in that path. This replaces the python search
path that much older versions needed.
* newlib isn't supported with oneApi so set TOOLCHAIN_HAS_NEWLIB to
OFF to match that.
* 2023.0.0 doesn't back llvm-nm, so use binutils version. This
is expected to be fixed in 2023.1.0 release so add a check to
handle either case.
* Update compiler flag check based on clang to also support
CMAKE_C_COMPILER_ID of "IntelLLVM" as that is how the oneApi toolchain
reports itself.
Signed-off-by: Kumar Gala <kumar.gala@intel.com>
LEGACY_INCLUDE_PATH has been defaulting to "n" and marked as deprecated
in both v3.2 and v3.3. Drop the option entirely for v3.4.
Signed-off-by: Fabio Baltieri <fabiobaltieri@google.com>
Move extra warning option from generic twister script into
compiler-dependent config files.
ARCMWDT compiler doesn't support extra warning options ex.
"-Wl,--fatal-warnings". To avoid build fails flag
"disable_warnings_as_errors" should be passed to twister.
This allows all warning messages and make atomatic test useles.
Signed-off-by: Nikolay Agishev <agishev@synopsys.com>
Deprecate old sparse support as Zephyr now provides a proper
infrastructure for SCA tools. Set ZEPHYR_SCA_VARIANT to sparse if user
is using deprecated way.
This allows to cleanup sparse code in various places and thus have a
cleaner build system.
Signed-off-by: Torsten Rasmussen <Torsten.Rasmussen@nordicsemi.no>
Add a compiler option to not merge globals. gen_kobject_list.py
is not capable of distinguish addresses of merged objects. The script
itself does not look wrong. The dward specification says that the
attribute DW_AT_location with opcode DW_OP_addr encodes a machine
address and whose size is the size of an address on the target machine
but for merged objects the address is longer, not clear why.
Disable global merge when userspace is enabled to avoid this problem.
Signed-off-by: Flavio Ceolin <flavio.ceolin@intel.com>
Disables allowing the python argparse library from automatically
shortening command line arguments, this prevents issues whereby
a new command is added and code that wrongly uses the shortened
command of an existing argument which is the same as the new
command being added will silently change script behaviour.
Signed-off-by: Jamie McCrae <jamie.mccrae@nordicsemi.no>
The build system currently links the toolchain-provided C++ standard
library even when the C++ support (`CONFIG_CPP`) is not enabled.
This commit updates the build system to link the toolchain-provided C++
standard library when the C++ support is enabled and a C++ library
implementation other than the Zephyr minimal C++ library is selected.
Signed-off-by: Stephanos Ioannidis <stephanos.ioannidis@nordicsemi.no>
This commit updates all deprecated `CONFIG_LIB_CPLUSPLUS` usages to:
* check if the Zephyr minimal C++ library is enabled using
`CONFIG_MINIMAL_LIBCPP` instead of relying on the
`CONFIG_LIB_CPLUSPLUS`-based inference.
* select `CONFIG_REQUIRES_FULL_LIBCPP` when there exists a component-
level C++ standard library dependency. This allows a component to
declare C++ standard library dependency without designating a
specific libray implementation.
* select the correct type of C++ standard library implementation to use
through one of the `CONFIG_LIBCPP_IMPLEMENTATION` choices.
Signed-off-by: Stephanos Ioannidis <stephanos.ioannidis@nordicsemi.no>
This commit updates all in-tree code to use `CONFIG_CPP_RTTI` instead
of `CONFIG_RTTI`, which is now deprecated.
Signed-off-by: Stephanos Ioannidis <stephanos.ioannidis@nordicsemi.no>
This commit updates all in-tree code to use `CONFIG_CPP_EXCEPTIONS`
instead of `CONFIG_EXCEPTIONS`, which is now deprecated.
Signed-off-by: Stephanos Ioannidis <stephanos.ioannidis@nordicsemi.no>
This commit updates all in-tree code to use `CONFIG_CPP` instead of
`CONFIG_CPLUSPLUS`, which is now deprecated.
Signed-off-by: Stephanos Ioannidis <stephanos.ioannidis@nordicsemi.no>
Add kobject_data prefix to kobject literals and group it close to
text area to avoid changing .text addresses in the final linkage.
Signed-off-by: Flavio Ceolin <flavio.ceolin@intel.com>
Generates XML file containing VIF policies by reading the device tree
using EDT.pickle generated during build
This script writes a subset of general and sink-pdo VIF policies in
output file
This script gets invoked during build if enabled through kconfig
The generated XML containing USB-C VIF policies could be used by
USB PD/Type-C analysers/testers to understand USB-C properties and
perform tests accordingly
Signed-off-by: Madhurima Paruchuri <mparuchuri@google.com>
This adds a choice of three different libc API buffer overflow detection
modes:
* None
* Compile-time
* Compile-time and Run-time
These correspond with the clang/gcc _FORTIFY_SOURCE modes (0/1/2).
_FORTIFY_SOURCE depends on compiler optimizations and require libc support
which the minimal C library doesn't include, so _FORTIFY_SOURCE is disabled
by default in those cases. Native tooling might also enable
_FORTIFY_SOURCE, so don't enable it by default in that case either.
Signed-off-by: Keith Packard <keithp@keithp.com>
GCC will compute expected sprintf (et al) return values internally and use
them in place of the actual return value. When the printf implementation
has reduced functionality, gcc may compute a different value.
For picolibc, this means disabling the optimization unless floating point
output is enabled.
Signed-off-by: Keith Packard <keithp@keithp.com>
This moves the test of compile flags set in environment from generic
toolchain to top-level ${ZEPHYR_BASE}/CMakeLists.txt.
In order to cleanup and make Zephyr CMake modules more generic then
the testing of specific compile flags and warn user is not really a
toolchain module responsible but a implementation detail of the project
itself as to how compile flags should be applied in a project.
Signed-off-by: Torsten Rasmussen <Torsten.Rasmussen@nordicsemi.no>
xcc-clang in based on clan-10 and does not support -fno-pic and
-fno-pie.
clang-10: warning: argument unused during compilation: '-fno-pic'
[-Wunused-command-line-argument]
[147/148] Linking C executable zephyr/zephyr.elf
clang-10: warning: argument unused during compilation: '-no-pie'
[-Wunused-command-line-argument]
Signed-off-by: Flavio Ceolin <flavio.ceolin@intel.com>
GCC 11 and above may generate a warning when dereferencing a constant
address pointer whose address is below the value specified by the
`min-pagesize` parameter.
This commit sets the `min-pagesize` parameter to 0 such that GCC never
generates the warnings for any constant address pointers.
For more details, refer to the GCC PR99578.
Signed-off-by: Stephanos Ioannidis <root@stephanos.io>
The app_smem linker script files were being generated by a python
script based on the contents of the cmake-generated
compile_commands.json file. But it forgot to include a dependency on
the input. It also has the misfeature of being generated at cmake
interpretation time with a real but empty stub.
This worked for quite a while, because the dependency on "kernel"
(it's not clear to me why this was there but I'm not brave enough to
remove it) forced the file to rebuild any time any of the C code in
the system changed, which was almost always guaranteed.
But then commit 0224f2c508 ("device.h:
rework the device_extern.h mechanism") came along and removed a
generated include from device.h and saved a bunch of build time. And
it also meant that no kernel code was being compiled anymore. So now,
a second build picked up that empty stub file and failed with linker
warnings about the orphan sections that didn't find a place in the
link.
But second builds after cmake re-invocations are an obscure edge case,
right? Well, no. It turns out that this is exactly what "twister -f"
does in CI, in order to get retries for test cases that fail
spuriously (mostly for unavoidable host timer-related glitches on SMP
platforms).
So the end result was that this problem showed itself as a silent
failure of the retry mechanism in twister/CI, making it look for all
the world like the timer tests had all suddenly gone flaky.
Fun times.
Signed-off-by: Andy Ross <andyross@google.com>
This commit adds a template for specifying the C/C++ compiler flag for
disabling the strict aliasing rule.
It also enables this flag globally because the Zephyr codebase does not
strictly adhere to the aliasing rules specified by the C/C++ standards
and the optimisation strategies that assume these rules may end up
generating invalid code.
For instance, GCC 11 and above tend to optimise more aggressively
assuming the strict adherence to the standard aliasing rules and may
generate invalid code, when compiling Zephyr with `-fstrict-aliasing`,
that results in various run-time failures.
Note that the footprint and performance ramifications of disabling the
strict aliasing rule are negligible.
Signed-off-by: Stephanos Ioannidis <root@stephanos.io>
Devicetree fixup files existed previous to the current stable Devicetree
API. While they served their purpose, they are no longer necessary nor
used in-tree. This patch drops support for this legacy feature.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
dts.fixup was a thing before dts_fixup.h existed. Remove the check as we
have already moved away from dts_fixup.h.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
