Both Clang [1] and (recently) GCC [2] support this flag to enable
additional codesize optimizations beyond -Os, possibly at the expense of
performance.
This tradeoff is worthwhile for some (and, Clang's -Oz seems to be
closer to GCC's -Os currently), so add a new abstraction for this flag
so users can select it as appropriate.
[1] https://clang.llvm.org/docs/CommandGuide/clang.html#cmdoption-O0
[2] https://gcc.gnu.org/onlinedocs/gcc/Optimize-Options.html#index-Oz
Signed-off-by: Jonathon Penix <jpenix@quicinc.com>
Use `do { ... } while (false)' instead of `do { ... } while (0)'.
Use comparisons with zero instead of implicitly testing integers.
Use comparisons with NULL instead of implicitly testing pointers.
Use comparisons with NUL instead of implicitly testing plain chars.
Use `bool' instead of `int' to represent Boolean values.
Use `while (true)' instead of `while (1)' to express infinite loops.
Signed-off-by: frei tycho <tfrei@baumer.com>
Create a Kconfig option for POSIX_MULTI_PROCESS Option Group
and mark it as experimental.
Currently, Zephyr has two functions that belong to this
Option Group:
- sleep()
- getpid()
Signed-off-by: Chris Friedt <cfriedt@tenstorrent.com>
This change deprecates CONFIG_POSIX_SIGNAL in favour of
the plural CONFIG_POSIX_SIGNALS, which maps directly to the
name of the standard POSIX Option Group POSIX_SIGNALS.
Additionally, mark signals as experimental.
Signed-off-by: Chris Friedt <cfriedt@tenstorrent.com>
This change deprecates CONFIG_POSIX_FS in favour of
CONFIG_POSIX_FILE_SYSTEM, which maps directly to the name of the
standard POSIX Option Group.
Signed-off-by: Chris Friedt <cfriedt@tenstorrent.com>
The POSIX_MAX_FDS option does not correspond to any standard
POSIX option. It was used to define the size of the file
descriptor table, which is by no means exclusively used by
POSIX (also net, fs, ...).
POSIX_MAX_FDS is being deprecated in order to ensure that
Zephyr's POSIX Kconfig variables correspond to those defined in
the specification, as of IEEE 1003.1-2017. Namely,
POSIX_OPEN_MAX. CONFIG_POSIX_MAX_OPEN_FILES is being deprecated
for the same reason.
To mitigate any possible layering violations, that option is
not user selectable. It tracks the newly added
CONFIG_ZVFS_OPEN_MAX option, which is native to Zephyr.
With this deprecation, we introduce the following Kconfig
options that map directly to standard POSIX Option Groups by
simply removing "CONFIG_":
* CONFIG_POSIX_DEVICE_IO
Similarly, with this deprecation, we introduce the following
Kconfig options that map directly to standard POSIX Options by
simply removing "CONFIG":
* CONFIG_POSIX_OPEN_MAX
In order to maintain parity with the current feature set, we
introduce the following Kconfig options.
* CONFIG_POSIX_DEVICE_IO_ALIAS_CLOSE
* CONFIG_POSIX_DEVICE_IO_ALIAS_OPEN
* CONFIG_POSIX_DEVICE_IO_ALIAS_READ
* CONFIG_POSIX_DEVICE_IO_ALIAS_WRITE
Gate open(), close(), read(), and write() via the
CONFIG_POSIX_DEVICE_IO Kconfig option and move
implementations into device_io.c, to be conformant with the
spec.
Lastly, stage function names for upcoming ZVFS work, to be
completed as part of the LTSv3 Roadmap (e.g. zvfs_open(), ..).
Signed-off-by: Chris Friedt <cfriedt@tenstorrent.com>
With this change, we collect implementations of functions that are
part of the POSIX_FD_MGMT Option Group into one compilation unit
(fd_mgmt.c) and provide a matching Kconfig symbol that is
reflective of the standard.
Additionally, provide Kconfig options for aliases of those functions
since Newlib-based toolchains internally use an underscore-prefixed
alias for many POSIX functions (in this case _lseek).
Prefix Zephyr-native implementations with zvfs_ to prepare for the
up-coming changes in that area.
Signed-off-by: Chris Friedt <cfriedt@tenstorrent.com>
The POSIX_CLOCK option does not correspond to any standard
option. It was used to active features of several distinct
POSIX Options and Option Groups, which complicated API and
application configuration as a result.
POSIX_CLOCK is being deprecated in order to ensure that Zephyr's
POSIX Kconfig variables correspond to those defined in the
specification, as of IEEE 1003.1-2017.
Additionally, CONFIG_TIMER is being deprecated because it does
not match the corresponding POSIX Option (_POSIX_TIMERS).
With this deprecation, we introduce the following Kconfig
options that map directly to standard POSIX Option Groups by
simply removing "CONFIG_":
* CONFIG_POSIX_TIMERS
Similarly, we introduce the following Kconfig options that
map directly to standard POSIX Options by simply removing
"CONFIG":
* CONFIG_POSIX_CLOCK_SELECTION
* CONFIG_POSIX_CPUTIME
* CONFIG_POSIX_DELAYTIMER_MAX
* CONFIG_POSIX_MONOTONIC_CLOCK
* CONFIG_POSIX_TIMEOUTS
* CONFIG_POSIX_TIMER_MAX
In order to maintain parity with the current feature set, we
introduce the following Kconfig options that map directly to
standard POSIX Option Groups by simply removing "CONFIG_":
* CONFIG_POSIX_MULTI_PROCESS - sleep()
Similarly, in order to maintain parity with the current feature
set, we introduce the following additional Kconfig options that
map directly to standard POSIX Options by simply removing
"CONFIG":
* CONFIG_XSI_SINGLE_PROCESS - gettimeofday()
Signed-off-by: Chris Friedt <cfriedt@tenstorrent.com>
Namespaced the generated headers with `zephyr` to prevent
potential conflict with other headers.
Introduce a temporary Kconfig `LEGACY_GENERATED_INCLUDE_PATH`
that is enabled by default. This allows the developers to
continue the use of the old include paths for the time being
until it is deprecated and eventually removed. The Kconfig will
generate a build-time warning message, similar to the
`CONFIG_TIMER_RANDOM_GENERATOR`.
Updated the includes path of in-tree sources accordingly.
Most of the changes here are scripted, check the PR for more
info.
Signed-off-by: Yong Cong Sin <ycsin@meta.com>
Zephyr minimal libc implements integer constant macros using internal
headers provided by GCC, for example INT64_C(x) is just a __INT64_C(x)
which is implemented by the GCC.
Clang compiler doesn't provide these definitions, so we define them
(__INT64_C() and INT64_C()) in zephyr/toolchain/llvm.h, but it looks
like INT64_C() definition doesn't come from llvm.h, but from stdlib.h
despite of checking if __GNUC__ is defined.
This is because Clang sometimes pretends to be GNU compiler:
$ x86_64-pc-linux-gnu-clang -dM -E - < /dev/null | grep "GNU"
#define __GNUC_MINOR__ 2
#define __GNUC_PATCHLEVEL__ 1
#define __GNUC_STDC_INLINE__ 1
#define __GNUC__ 4
Let's keep the integer constant macros in stdlib.h and their
actual implementation in llvm.h. Also define these macros if __clang__
is defined, just in case.
Signed-off-by: Patryk Duda <patrykd@google.com>
Allow Picolibc to get build from module, if the toolchain does not
include a bundled picolibc.
This should always be possible, except for native applications and
if libcpp is required.
Signed-off-by: David Schneider <david.schneider@chargepoint.com>
This aligns abort() and exit() definitions with other libc.
Without 'noreturn' attribute, compilers have to assume that we will
return from these functions which can lead to surprising errors like
'error: non-void function does not return a value'.
Signed-off-by: Patryk Duda <patrykd@google.com>
Let's introduce a new option so components which cannot
be used with the native libC can select it.
And at this point let's already have POSIX_API select it
instead of being used in the libC choice default.
Signed-off-by: Alberto Escolar Piedras <alberto.escolar.piedras@nordicsemi.no>
Support getting and setting POSIX environment variables.
Additionally, the thread-safe BSD variant getenv_r() is
provided.
environ, getenv(), setenv(), and unsetenv() are required by
the POSIX_SINGLE_PROCESS Option Group as detailed in
Section E.1 of IEEE-1003.1-2017.
The POSIX_SINGLE_PROCESS Option Group is required for PSE51,
PSE52, PSE53, and PSE54 conformance, and is otherwise mandatory
for any POSIX conforming system as per Section A.2.1.3 of
IEEE-1003-1.2017.
Signed-off-by: Christopher Friedt <cfriedt@meta.com>
It turns out that currently LTO is enabled only for the kernel.
This commit updates it to enable it for the whole application
and adds additional LTO exclusions required for the standard
C libraries to build and link properly.
Signed-off-by: Radosław Koppel <radoslaw.koppel@nordicsemi.no>
gmtime_r() has been in the minimal libc for years, however it was not
added to expcetions due to an overlook. In order to do this however, it
has to be moved first to the common libc area, so that it's available
to any libc that may not implement it.
Signed-off-by: Robert Lubos <robert.lubos@nordicsemi.no>
Newlib doesn't have Zephyr support, so we need to define these functions
when the application doesn't ask for the right level of POSIX support.
Signed-off-by: Keith Packard <keithp@keithp.com>
This was necessary to get Picolibc to expose the whole Zephyr C library
API, but current versions of the SDK use a version of Picolibc with
built-in Zephyr support.
Signed-off-by: Keith Packard <keithp@keithp.com>
Since pthread_once() is both the initializer and executor of
pthread_once_t, it can have maximally two states. Since the
implementation in Zephyr previously aimed to maximize libc
compatibility, we opted to use the definition of pthread_once_t
from newlib, which is a structure with two ints.
It does not make sense to use 64 bits to manage 2 possible
states. The control for that should effectively be a bool.
We maintain compatibility with newlib by asserting (at build
time), that newlib's pthread_once_t is larger than Zephyr's
new struct pthread_once (which just contains a bool).
This allows us to delete the non-standard pthread_key.h
header file (finally).
Reuse the pthread_pool_lock in order to synchronize the related
init function (so that it is only called maximally once from any
thread). The spinlock is only used to test the state and the
init function is not called with the spinlock held.
The alternative was to use an atomic inside of
struct pthread_once. But again, that could be up to 64-bits with
Zephyr's atomics implementation.
Ideally we would use C11 generics or something to support atomics
on 8, 16, 32, and 64-bit primitives.
Lastly, also update declarations for C11 threads as they mostly
mirror our pthread implementation.
This needed to be done as a single commit in order to ensure
continuity of build.
Signed-off-by: Christopher Friedt <cfriedt@meta.com>
The string "" is of type '(const char *)', so add a cast over to
'(char *)' to clean up source code.
Signed-off-by: Florian La Roche <Florian.LaRoche@gmail.com>
Picolibc's 'minimal' printf mode reduces functionality and size even more
than the 'integer' mode. Use this where memory is at a premium and where
the application knows that it does not require exact printf semantics.
1.8.5 adds two more printf variants, 'long long' and 'minimal'. The 'long
long' variant is the same as the 'integer' variant but with long long
support enabled. The 'minimal' variant reduces functionality and size even
more than the 'integer' mode. Applications can use this where memory is at
a premium and where the application does not require exact printf
semantics.
With these two added variants, the SDK has enough options so that all of
the cbprintf modes can be supported with the pre-compiled bits:
1. CBPRINTF_NANO - picolibc's 'minimal' variant
2. CBPRINTF_REDUCED_INTEGRAL - picolibc's 'integer' variant
3. CBPRINTF_FULL_INTEGRAL - picolibc's 'long long' variant
4. CBPRINTF_FB_SUPPORT - picolibc's 'double' variant
This patch makes the cbprintf Kconfig values drive the default picolibc
variant, disables picolibc variants not capable of supporting the required
cbprintf level, but allows applications to select more functionality in
picolibc than cbprintf requires.
Note that this depends on the SDK including picolibc 1.8.5. Without that,
selecting the 'minimal' or 'long long' variant in Zephyr will end up with
the default variant from picolibc, which is the full version with floating
point support. When using the module things will work as specified.
Signed-off-by: Keith Packard <keithp@keithp.com>
This option in picolibc switches the assert macro between a chatty version
and one which provides no information at all. This latter mode avoids
placing the associated strings in memory.
The Zephyr option is PICOLIBC_ASSERT_VERBOSE and it is disable by default.
Signed-off-by: Keith Packard <keithp@keithp.com>
This moves the k_* memory management functions from sys/ into
kernel/ includes, as there are kernel public APIs. The z_*
functions are further separated into the kernel internal
header directory.
Also made a quick change to doxygen to group sys_mem_* into
the OS Memory Management group so they will appear in doc.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
This change capitalizes on newly added support for dynamic
thread stacks and the existing pthread support to provide
an implementation of the ISO C11 `<threads.h>` API.
Signed-off-by: Christopher Friedt <cfriedt@meta.com>
This prevents the compiler from optimizing calloc into an
infinite recursive call.
For example a call to malloc + memset zero at GCC -O2 will be
replaced by a call to calloc. This causes infinite recursion
if the function being implemented *is* calloc.
fno-builtin-malloc forces the compiler to avoid this optimization.
Signed-off-by: Grant Ramsay <gramsay@enphaseenergy.com>
The picolibc module can be built without thread local storage support if
desired. Allow that by using 'imply' instead of 'select'. However, when
using the toolchain picolibc, we assume that TLS will be enabled wherever
supported, so make sure we match by adding a 'select' for this case.
Signed-off-by: Keith Packard <keithp@keithp.com>
Move the syscall_handler.h header, used internally only to a dedicated
internal folder that should not be used outside of Zephyr.
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
Instead of making applications use C library specific settings to enable
floating point support in printf, provide this indirect symbol which then
detects which C library is in use and selects the correct configuration for
each.
Signed-off-by: Keith Packard <keithp@keithp.com>
When Zephyr is compiled using LLVM toolchain, we don't need to link with
libgcc to resolve libc dependencies. With this patch, the trick will be
applied only when the GNU compiler is used. Otherwise, we will just link
libc, which works for LLVM.
Signed-off-by: Patryk Duda <pdk@semihalf.com>
Defaulting to picolibc when selecting GETOPT is a bit nicer for users
in some cases, but not required.
But too many things require GETOPT (for ex. some SHELL configurations)
in combinations with native posix drivers which do not support
yet embedded libcs (for ex. the native USB driver)
(see https://github.com/zephyrproject-rtos/zephyr/issues/60096 )
Which leads to configurations in those cases which cannot be built.
Keep defaulting to the external libC in this case.
This reverts the getop part of this commit:
5f8057e262
Signed-off-by: Alberto Escolar Piedras <alberto.escolar.piedras@nordicsemi.no>
When building with the POSIX_API we cannot use the host libC.
When using GETOPT it is similarly quite difficult for users
to use the host libC.
So to make it easier for users, let's just default to PICOBLIC
in those cases,
while we continue defaulting to the host library in other cases
so users can use the Linux APIs for whatever test functionality
they want.
Signed-off-by: Alberto Escolar Piedras <alberto.escolar.piedras@nordicsemi.no>
Leave the malloc partition so that it only contains the heap itself; this
lets the initialization code adjust the address range when configuring the
arena at startup.
Signed-off-by: Keith Packard <keithp@keithp.com>
Instead of adding every possible subsystem which places variables in the C
library memory partition in libc-hooks.h, place those conditions in the
related Kconfig files and simplify the libc-hooks.h to just looking at
CONFIG_NEED_LIBC_MEM_PARTITION.
Signed-off-by: Keith Packard <keithp@keithp.com>
Picolibc dependencies limit ability to use third party minimal
implementations of CPP when enablng PICOLIBC_USE_MODULE.
Signed-off-by: Al Semjonovs <asemjonovs@google.com>
Fixes#62677 where printf defaults to using armclang's semihosting backed
implementation of printf.
Signed-off-by: Wilfried Chauveau <wilfried.chauveau@arm.com>
The original idea of z_current_get() was to be the counterpart
of k_current_get() when thread local variable for current has
not been initialized if TLS is enabled, otherwise they are
the same function. Now since z_current_get() is being used
outside of core kernel, rename it under kernel namespace so
other subsystem can conceptually use them too.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
When multithreading is disabled, the whole mutex infrastructure isn't
available. The common malloc code wasn't checking for this case which
caused build failures.
Signed-off-by: Keith Packard <keithp@keithp.com>
Tested-by: Detlev Zundel dzu@member.fsf.org
