A speciality of the RTEMS build system was the make preinstall step. It
copied header files from arbitrary locations into the build tree. The
header files were included via the -Bsome/build/tree/path GCC command
line option.
This has at least seven problems:
* The make preinstall step itself needs time and disk space.
* Errors in header files show up in the build tree copy. This makes it
hard for editors to open the right file to fix the error.
* There is no clear relationship between source and build tree header
files. This makes an audit of the build process difficult.
* The visibility of all header files in the build tree makes it
difficult to enforce API barriers. For example it is discouraged to
use BSP-specifics in the cpukit.
* An introduction of a new build system is difficult.
* Include paths specified by the -B option are system headers. This
may suppress warnings.
* The parallel build had sporadic failures on some hosts.
This patch removes the make preinstall step. All installed header
files are moved to dedicated include directories in the source tree.
Let @RTEMS_CPU@ be the target architecture, e.g. arm, powerpc, sparc,
etc. Let @RTEMS_BSP_FAMILIY@ be a BSP family base directory, e.g.
erc32, imx, qoriq, etc.
The new cpukit include directories are:
* cpukit/include
* cpukit/score/cpu/@RTEMS_CPU@/include
* cpukit/libnetworking
The new BSP include directories are:
* bsps/include
* bsps/@RTEMS_CPU@/include
* bsps/@RTEMS_CPU@/@RTEMS_BSP_FAMILIY@/include
There are build tree include directories for generated files.
The include directory order favours the most general header file, e.g.
it is not possible to override general header files via the include path
order.
The "bootstrap -p" option was removed. The new "bootstrap -H" option
should be used to regenerate the "headers.am" files.
Update #3254.
Use a self-contained recursive mutex for API_Mutex_Control. The API
mutexes are protected against asynchronous thread cancellation.
Add dedicated mutexes for libatomic and TOD.
Close#2629.
Close#2630.
The previous rtems_panic() implementation was quite heavy weight. It
depended on _exit() which calls the global destructors. It used
fprintf(stderr, ...) for output which depends on an initialized console
device and the complex fprintf().
Introduce a new fatal source RTEMS_FATAL_SOURCE_PANIC for rtems_panic()
and output via vprintk().
Update #3244.
Provide functions to get the version string, major, minor and revision
numbers and the version control identifer that is a unique tag for
the version control system.
Update #3199.
The EDF SMP scheduler supports simple thread processor affinities
(see #3059) with a small run-time overhead. The current default SMP
scheduler lacks support for thread processor affinities at all. The EDF
SMP scheduler offers a good feature set for most applications. So, use
it by default. Run-time libraries like libgomp, MTAPI, work stealing
schedulers, language interpreters (e.g. Erlang virtual machine),
maintainence of per-processor data (e.g. Universal Memory Allocator
(UMA)), etc. use a one-to-one thread processor affinity for example.
Update #3063.
Do not use the processor count to determine if SMP is enabled. Instead
use a dedicated configuration option. Enable SMP by default in SMP
configurations.
Add CONFIGURE_DISABLE_SMP_CONFIGURATION to all test which would fail
otherwise.
Update #3001.
Enable the SMP configuration by default in case SMP is enabled. Add
configuration option CONFIGURE_DISABLE_SMP_CONFIGURATION to disable it
explicitly.
Add CONFIGURE_DISABLE_SMP_CONFIGURATION to all test which would fail
otherwise.
Update #3001.
Add POSIX shared memory manager (Shm). Includes a hook-based
approach for the backing memory storage that defaults to the
Workspace, and a test is provided using the heap. A test is
also provided for the basic use of mmap'ing a shared memory
object. This test currently fails at the mmap stage due to
no support for mmap.
The fatal is internal indicator is redundant since the fatal source and
error code uniquely identify a fatal error. Keep the fatal user
extension is internal parameter for backward compatibility and set it to
false always.
Update #2825.
Turn pthread_spinlock_t into a self-contained object. On uni-processor
configurations, interrupts are disabled in the lock/trylock operations
and the previous interrupt status is restored in the corresponding
unlock operations. On SMP configurations, a ticket lock is a acquired
and released in addition.
The self-contained pthread_spinlock_t object is defined by Newlib in
<sys/_pthreadtypes.h>.
typedef struct {
struct _Ticket_lock_Control _lock;
__uint32_t _interrupt_state;
} pthread_spinlock_t;
This implementation is simple and efficient. However, this test case of
the Linux Test Project would fail due to call of printf() and sleep()
during spin lock ownership:
https://github.com/linux-test-project/ltp/blob/master/testcases/open_posix_testsuite/conformance/interfaces/pthread_spin_lock/1-2.c
There is only limited support for profiling on SMP configurations.
Delete CORE spinlock implementation.
Update #2674.
Rename _Scheduler_Assignments into _Scheduler_Initial_assignments to
make it clear that they may not reflect the run-time scheduler
assignment.
Update #2797.
The _RBTree_Find() is no longer used in the score. Move it to sapi and
make it rtems_rbtree_find(). Move corresponding types and support
functions to sapi.
This change adds rtems_printf and related functions and wraps the
RTEMS print plugin support into a user API. All references to the
plugin are removed and replaced with the rtems_printer interface.
Printk and related functions are made to return a valid number of
characters formatted and output.
The function attribute to check printf functions has been added
to rtems_printf and printk. No changes to remove warrnings are part
of this patch set.
The testsuite has been moved over to the rtems_printer. The testsuite
has a mix of rtems_printer access and direct print control via the
tmacros.h header file. The support for begink/endk has been removed
as it served no purpose and only confused the code base. The testsuite
has not been refactored to use rtems_printf. This is future work.
The priority values are only valid within a scheduler instance. Thus,
the maximum priority value must be defined per scheduler instance. The
first scheduler instance defines PRIORITY_MAXIMUM. This implies that
RTEMS_MAXIMUM_PRIORITY and POSIX_SCHEDULER_MAXIMUM_PRIORITY are only
valid for threads of the first scheduler instance. Further
API/implementation changes are necessary to fix this.
Update #2556.
