Add rtems_interrupt_server_destroy().
Before this patch, the only way to create interrupt servers was
rtems_interrupt_server_initialize(). This function creates the default
interrupt server and in SMP configurations additional interrupt servers
for the additional processors. The interrupt server is heavily used by
libbsd. This includes the epoch based reclamation which performs time
consuming resource and memory deallocation work. This does not work well
with time critical services, for example an UART over SPI or I2C. One
approach to address this problem is to allow the application to create
custom interrupt servers with the right priority and task properties.
The interrupt server API accounted for this, however, it was not
implemented before this patch.
Close#4034.
If we interrupt a thread dispatch critical section (thread dispatch
disable level != ISR nest level), then we should not do the fast idle
mode since this may delay an ongoing system call forever.
CLOCK_DRIVER_ISRS_PER_TICK is the configuration define,
CLOCK_DRIVER_ISRS_PER_TICK_VALUE is the actual value of ISRS per clock
tick, therefore use this one to reset the Clock_driver_isrs after each
tick.
This patch will allow the user to pass a function to calculate
the baud divisor.
This will allow for more flexibility, since for some BSPs
like raspberrypi, the calculation of baud divisor is different
from what is in the current driver.
Use RTEMS_SYSINIT_ORDER_LAST_BUT_5 instead of RTEMS_SYSINIT_ORDER_LAST
to allow applications and support functions to place system
initialization handlers behind the standard handlers.
Update #3838.
The work area initialization was done by the BSP through
bsp_work_area_initialize(). This approach predated the system
initialization through the system initialization linker set. The
workspace and C program heap were unconditionally initialized. The aim
is to support RTEMS application configurations which do not need the
workspace and C program heap. In these configurations, the workspace
and C prgram heap should not get initialized.
Change all bsp_work_area_initialize() to implement _Memory_Get()
instead. Move the dirty memory, sbrk(), per-CPU data, workspace, and
malloc() heap initialization into separate system initialization steps.
This makes it also easier to test the individual initialization steps.
This change adds a dependency to _Heap_Extend() to all BSPs. This
dependency will be removed in a follow up change.
Update #3838.
Resource leak identified by Coverity (CID 1456675). The value
of instances is leaked in case some but not all irq servers are
created. It should be stored in bsp_interrupt_server_instances.
The use case for this function is the libbsd. In FreeBSD, the interrupt
setup and binding to a processor is done in two steps. Message
based interrupts like PCIe MSI and MSI-X interrupts can be implemented
through interrupt server entries. They are setup at the default
interrupt server and may optionally move to an interrupt server bound to
a specific processor.
Move resposibility to disable thread dispatching to the caller of
_SMP_Multicast_action(). Using an interrupt disable for this purpose is
questionable.
Rename _SMP_Get_processor_count() in _SMP_Get_processor_maximum() to be
in line with the API level rtems_scheduler_get_processor_maximum().
Update #3732.
Add rtems_scheduler_get_processor_maximum() as a replacement for
rtems_get_processor_count(). The rtems_get_processor_count() is a bit
orphaned. Adopt it by the Scheduler Manager. The count is also
misleading, since the processor set may have gaps and the actual count
of online processors may be less than the value returned by
rtems_get_processor_count().
Update #3732.
- Use CamelCase as it is not used in our C code. Enables simple search and
replace.
- Prefix with "RTEMS" to aid deployment and integration. It aids
searching and sorting.
Update #3706.
In order to simplify the use of CPU counter values it is beneficial to
have monotonic increasing values within the range of the CPU counter
ticks data type, e.g. 32-bit unsigned integer. This eases the use of
CPU counter timestamps in external tools which do not know the details
of the CPU counter hardware. The CPU counter is the fastest way to get
a time on an RTEMS system.
Such a CPU counter may be also used as the timecounter. Use it on SPARC
for this purpose to simplify the clock drivers.
Update #3456.
The following rtems_cache_*_processor_set() cache manager API functions
are exotic, complex, very hard to use correctly, not used in the RTEMS
code base, and apparently unused by applications.
Close#3622.
The precision clock synthesizer support broke the driver on the QorIQ
P1020. On this device the Alternate Function Register is accessed with
DLAB == 1 instead of the FIFO Control Register (FCR). Restructure the
code to account for this.
Add and use <machine/rtems-bsd-kernel-space.h> and
<machine/rtems-bsd-user-space.h> similar to the libbsd to avoid command
line defines and defines scattered throught the code base.
Simplify cpukit/libnetworking/Makefile.am.
Update #3375.
