We don't need MWDT libc if we are using minimal one provided by
Zephyr. So let's avoid linking mwdt libc if CONFIG_MINIMAL_LIBC
is enabled.
Signed-off-by: Eugeniy Paltsev <Eugeniy.Paltsev@synopsys.com>
Signed-off-by: Evgeniy Paltsev <PaltsevEvgeniy@gmail.com>
MetaWare hostlink is a library of system calls for debugging.
We don't use in case of Zephyr, however we provide real system
calls implementations. Let's pass -Hhostlib= to MWDT linker to
eliminate the MetaWare hostlink library from linkage.
Signed-off-by: Eugeniy Paltsev <Eugeniy.Paltsev@synopsys.com>
Signed-off-by: Evgeniy Paltsev <PaltsevEvgeniy@gmail.com>
The file <board>.dts_compiled was made obsolete in 8d8b06978c.
<board>.dts_compiled contains only a reference to zephyr.dts.
Users interested in devicetree should be familiar with zephyr.dts,
therefore it's time to remove the unused <board>.dts_compiled.
Signed-off-by: Torsten Rasmussen <Torsten.Rasmussen@nordicsemi.no>
add a config CONFIG_QEMU_UEFI_BOOT to indicate whether
the qemu will use UEFI bootable method;
add a new test "sample.basic.helloworld.uefi" to verify
UEFT bootable method on qemu_x86_64 platform.
Signed-off-by: Chen Peng1 <peng1.chen@intel.com>
I saw a user wondering what this is for and and where it comes from.
Provide a comment header with some explanation and a pointer to where
to find out more.
Signed-off-by: Martí Bolívar <marti.bolivar@nordicsemi.no>
The constructors of static objects are stored in ".ctors"
section. In case of MWDT toolchain we have incompatible
".ctors" section format with GNU toolchain. So let's use
initialization code provided by MWDT instead of Zephyr one
in case of MWDT toolchain usage.
As it is done for GNU toolchain We call constructors of
static objects but we don't call destructors for them.
Signed-off-by: Eugeniy Paltsev <Eugeniy.Paltsev@synopsys.com>
Signed-off-by: Evgeniy Paltsev <PaltsevEvgeniy@gmail.com>
In order to use the newlib retagetable locking interface (for thread
safety), which requires the newlib multi-threading feature to be
enabled, the Zephyr SDK 0.13.0 or above must be used.
Signed-off-by: Stephanos Ioannidis <root@stephanos.io>
Current location options for linker source files includes init and
noinit ram data, but only a noinit ram section. This makes it impossible
for application code to define an initialized RAM output section,
such as with the Z_ITERABLE_SECTION_RAM() helpers.
Adding a DATA_SECTIONS linker source option for this use case.
Signed-off-by: Fabio Baltieri <fabiobaltieri@google.com>
add .S file extension suffix into CMAKE_ASM_SOURCE_FILE_EXTENSIONS,
because clang from OneApi can't recongnize them as asm files on
windows, then they won't be added into build system.
Signed-off-by: Chen Peng1 <peng1.chen@intel.com>
The icx compiler of oneApi will invoke clang-cl on windows,
this is not supported in zephyr now, so change to use
clang directly.
And the clang from oneApi can't recognize those cross
compiling target variables of cmake, such as
CMAKE_C_COMPILER_TARGET, so used other cmake functions
to pass them to clang.
Signed-off-by: Chen Peng1 <peng1.chen@intel.com>
The compiler option "-fdiagnostics-color=always" is not known to XCC
(which is based on really old GCC). So don't enable color
diagnostic output when building with XCC.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
Add auto-generated kconfig header to linker script pre-processing
via -imacros flag. This permits link scripts supplied via
zephyr_linker_sources() to use CONFIG_* variables.
Signed-off-by: David Palchak <palchak@google.com>
Add some more example to the description of zephyr_library_amend().
This should help users to get input of the extra possibilities that this
function provides.
See: #35770
Signed-off-by: Torsten Rasmussen <Torsten.Rasmussen@nordicsemi.no>
Signed-off-by: Andrzej Głąbek <andrzej.glabek@nordicsemi.no>
-Hcl option instructs linker to use MetaWare C Compactlib.
According to MWDT documentation "Compactlib is not thread-safe"
So, let's use MetaWare C library instead which is thread-safe.
Signed-off-by: Evgeniy Paltsev <PaltsevEvgeniy@gmail.com>
With respect to the mapping between the CPU variant
and the FPU we add an entry for Cortex-M v8-M CPU
variants without DSP. This should cover the case
of a Mainline Cortex-M which implements the Floating
Point extension but not the DSP extension.
Signed-off-by: Ioannis Glaropoulos <Ioannis.Glaropoulos@nordicsemi.no>
When the host toolchain is used, `-nostdlib` option is not specified
and therefore all default libraries, including the libgcc and its
components, are automatically linked -- so it is not necessary to
manually link them here.
Signed-off-by: Stephanos Ioannidis <root@stephanos.io>
The host toolchain makes use of the host toolchain libraries (i.e.
`-nostdlib` is not specified), so it is not necessary to detect the
toolchain libgcc path and specify one manually.
Signed-off-by: Stephanos Ioannidis <root@stephanos.io>
This commit updates the build system to always detect the libgcc path
for the host platforms regardless of the target architecture.
The native_posix_64 target previously used the x32 ABI (`-mx32`) and
the multilib for this ABI was not commonly available in the x86-64 host
toolchains.
That is no longer the case because native_posix_64 had been updated to
use the x86-64 ABI (`-m64`).
Signed-off-by: Stephanos Ioannidis <root@stephanos.io>
This commit removes the `NOT no_libgcc` condition for the automatic
libgcc path detection.
The `no_libgcc` variable was previously set in this file to exclude
libgcc detection for the x86-64 targets because the x86-64 arch used
x32 ABI (`-mx32`) and the multilib for this ABI was not commonly
available. This is no longer the case (x86-64 targets now use `-m64`)
and libgcc must always be available.
Signed-off-by: Stephanos Ioannidis <root@stephanos.io>
Fixes: #36185
CMake >=3.20 now supports relative paths in Ninja depfiles.
As Zephyr still supports the use of CMake 3.13, and the current use of
DEPFILE in linker/ld/target.cmake is already absolute path, then the
safest choice is to use old behavior.
This can be changed when Zephyr increases minimum required CMake to be
3.20 or newer.
Signed-off-by: Torsten Rasmussen <Torsten.Rasmussen@nordicsemi.no>
Always source the Zephyr base soc and shield (board root) Kconfigs
directly from Kconfig instead of generated Kconfig files.
This has the benefit that it is no longer necessary to generate Kconfig
files to source SoC root and shield (board root) when no custom roots
are provided.
Also this cleans up the doc/CMakeLists.txt and ensures that the
doc/CMakeLists.txt is not getting out of sync with the Kconfig.cmake.
Signed-off-by: Torsten Rasmussen <Torsten.Rasmussen@nordicsemi.no>
With this change, we can put contents into rom_start section
by calling zephyr_linker_sources(ROM_START ...)
Signed-off-by: Dino Li <Dino.Li@ite.com.tw>
Co-authored-by: Torsten Tejlmand Rasmussen
Change-Id: If1169423b013d3e4df52d91cdb2fbdddc3bace7b
GCC 11 defaults to output DWARF version 5 which pyelftools cannot
currently parse. So keep output at version 4 for the time being.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
Fixes: #35890
The current implementation of Zephyr build configuration CMake package
only allows the build configuration package to be placed inside a
Zephyr workspace.
This commit extends the usability by allowing users to locate the
Zephyr build configuration CMake package outside the Zephyr workspace
and then refer to the package using
`-DZephyrBuildConfiguration_ROOT=<path>`
`set(ZephyrBuildConfiguration_ROOT <path>)`
This allows users greater flexibility in their workspace layouts.
Signed-off-by: Torsten Rasmussen <Torsten.Rasmussen@nordicsemi.no>
Fixes: #19582
When `find_package(Zephyr)` completes then all boilerplate code has been
processed and all Zephyr libraries has been placed inside the
whole archive flags.
Also all libs dependencies has been processed.
This is indicated by the presence of the zephyr_prebuilt target.
Thus, warn the user if `zephyr_library()` / `zephyr_library_named()`
is called in app-mode.
Signed-off-by: Torsten Rasmussen <Torsten.Rasmussen@nordicsemi.no>
Add a tree node to group files under WORKSPACE so that
they won't be shown with full path. The WORKSPACE is
usually the same as WEST_TOPDIR unless ZEPHYR_WORKSPACE
is defined during build.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
Fixes part of: #32448
This commit updates the CMake CMAKE_CXX_LINK_EXECUTABLE to include
crtbegin.o and crtend.o at the right locations when linking with gcc.
It also updates linker scripts to ensure proper location of the
exception header frame sections.
This ensure proper handling of exceptions for those architectures
- x86
- xtensa
- riscv32
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
Signed-off-by: Torsten Rasmussen <Torsten.Rasmussen@nordicsemi.no>
After the ARM/ARM64 split the SDK 0.12.4 is needed to be able to compile
the ARM64 arch code. Bail out if an old SDK is detected at compile time.
Also change the minimal supported Zephyr SDK to v0.12, as agreed in
Toolchain WG. Main reason is added compiler architectures and important
inclusion of fixes, for example qemu and arm fixes for veneers (TFM)
Signed-off-by: Carlo Caione <ccaione@baylibre.com>
Due to the fact that we use -mcpu=hs as a default for ARCv2 HS
the compiler doesn't generate multiply/divide/mac/64bit memory
operations instructions.
Fix that by using -mcpu=archs as a default for ARCv2 HS which is
fits for all existing boards with ARCv2 HS CPU.
Signed-off-by: Eugeniy Paltsev <Eugeniy.Paltsev@synopsys.com>
Signed-off-by: Evgeniy Paltsev <PaltsevEvgeniy@gmail.com>
Fixes: #32237
When building for native_posix, then host tools are used.
This means that gcc will link using `/usr/bin/ld` per default.
If ld points to lld, then linking will fail.
This commit will first look for ld.bfd, and if found then use
-fuse-ld=bfd for linking. If ld.bfd is not found, then ld is used as
fallback as that will be assumed to be the best working candidate.
Signed-off-by: Torsten Rasmussen <Torsten.Rasmussen@nordicsemi.no>
There is a comment which needed to be rephrased for inclusive
language. This commit replaces it with the recommended
replacement.
Signed-off-by: Jennifer Williams <jennifer.m.williams@intel.com>
Adds no-threadsafe-statics to compiler flags for gcc and arcmwdt.
The flag is required to compile the samples - the module can compile
without it.
Signed-off-by: Lauren Murphy <lauren.murphy@intel.com>
Follow-up: #34842
This is a follow-up commit that ensures relevant version information
regarding the build is located together as discussed in #34842.
Signed-off-by: Torsten Rasmussen <Torsten.Rasmussen@nordicsemi.no>
If user is specifying `-DBUILD_VERSION=<value>` then the proper way to
test is using `NOT DEFINED BUILD_VERSION`.
The difference between `NOT DEFINED BUILD_VERSION` and
`NOT BUILD_VERSION` is that a user specifying `-DBUILD_VERSION=0` will
result in `NOT BUILD_VERSION` becoming true, whereas
`NOT DEFINED BUILD_VERSION` would be false, which is the correct
behavior here, as the user actually provided a specific BUILD_VERSION.
Signed-off-by: Torsten Rasmussen <Torsten.Rasmussen@nordicsemi.no>
Do basic preparations for building code for ARCv3 HS6x
* add ISA_ARCV3 and CPU_HS6X config options
* add off_t type support for __ARC64__
* use elf64-littlearc format for linking
* use arc64 mcpu for CPU_HS6X
Signed-off-by: Eugeniy Paltsev <Eugeniy.Paltsev@synopsys.com>
Signed-off-by: Evgeniy Paltsev <PaltsevEvgeniy@gmail.com>
As of today we check against CPU_ARCHS config option which may be
related to different HS CPUs (i.e. ARCv2 HS, ARCv3 HS, etc...).
Let's check against exact HS CPU type (CPU_HS3X) as we do EM
processors.
Signed-off-by: Eugeniy Paltsev <Eugeniy.Paltsev@synopsys.com>
Signed-off-by: Evgeniy Paltsev <PaltsevEvgeniy@gmail.com>
In an ideal world, good CIs make it very clear what exact git versions
are getting built. However:
- Zephyr is (re-)used in many projects and they cannot all be expected
to have ideal CI.
- CI with multiple git repos is complex (#34713) which makes the world
an even less ideal place: much more chance for some git versions to be
missing.
- Many developers don't realize that Github and other CIs do not test
pull/12345/head but the moving target pull/12345/merge instead. While
not resolving pull/12345/merge completely (maybe another day), this
commit provides at least evidence that pull/12345/head is NOT the
commit tested.
So the addition of the following line in the logs is a very small price
to pay that can save enormous amounts of time when trying to understand
some obscure build failures.
-- BUILD_VERSION=zephyr-v2.5.0-2957-g6230b5bb66bc
Note this obviously does not provide any git information when
BUILD_VERSION is overriden but it does not hurt either in this case:
knowing BUILD_VERSION was overriden is also useful.
The "BUILD_VERSION is left undefined" message was just wrong in
the (unlikely) case `git describe` printed something while also
failing. Remove it; it's so much simpler to just print $BUILD_VERSION
and give the direct, unfiltered information.
Note this simplification is also a partial revert of 1b80f00f56
which threw the entire git warnings "baby" with some obscure duplicate
1.13.0 "bathwater" that is not relevant any more and that I guess barely
anyone noticed even at the time.
Signed-off-by: Marc Herbert <marc.herbert@intel.com>
Fixes: #34368
Now added WEST_PYTHON to list of preferred pythons to check.
WEST_PYTHON is added after PYTHON_PREFER so that a user may still
request a specific python is used by supplying
`-DPYTHON_PREFER=<python>` even when building using `west build`.
Signed-off-by: Torsten Rasmussen <Torsten.Rasmussen@nordicsemi.no>
Today, there is a build target is added for each runner: flash, debug,
debugserver, attach.
And those runners will have a dependency to Zephyr logical target that
is built before invoking `west <runner>`.
This design has some flaws, mainly that additional dependencies directly
on the target will not be built when running `west <runner>` directly.
That generator expressions cannot be used for the DEPENDS argument.
Instead, the build target `<runner>` will not have any dependencies, and
will raise a build error if a dependency is added to the target.
Due to how `add_dependencies()` work, this must be done as a build time
check, and not configure time check.
`west <runner>` will invoke a build before executing the runner, and
this way ensure the build target is up-to-date, which again removes the
need for a dedicated `west_<runner>_target`.
It also minimizes the risk of developer errors, as developers no longer
need to consider the need for adding additional dependencies.
If a custom target is part of the default `all` build, then it's ensured
to be up-to-date.
Fixes: Issue reported on slack.
Signed-off-by: Torsten Rasmussen <Torsten.Rasmussen@nordicsemi.no>
Some targets require *strip* command. This command was not
being set when selecting llvm toolchain.
Signed-off-by: Flavio Ceolin <flavio.ceolin@intel.com>
When compiler results are piped through a non-terminal (e.g. ninja)
the compiler disables colour diagnostics. Using `-fdiagnostics-color`
forces the compiler to enable colour output. This flag is set for
clang and gcc when `ZEPHYR_BUILD_COLOUR_DIAGNOSTIC` environment
variable is set when a clean build is started.
Signed-off-by: Arvin Farahmand <arvinf@ip-logix.com>
The parameter can be used to limit the ram and rom reports to a
certain depth. The resulting graphs with depths of e.g. 3 or 4 are
visually easier to grasp.
Signed-off-by: Detlev Zundel <dzu@member.fsf.org>
