Clustered/partitioned scheduling helps to control the worst-case
latencies in the system. The goal is to reduce the amount of shared
state in the system and thus prevention of lock contention. Modern
multi-processor systems tend to have several layers of data and
instruction caches. With clustered/partitioned scheduling it is
possible to honour the cache topology of a system and thus avoid
expensive cache synchronization traffic.
We have clustered scheduling in case the set of processors of a system
is partitioned into non-empty pairwise-disjoint subsets. These subsets
are called clusters. Clusters with a cardinality of one are partitions.
Each cluster is owned by exactly one scheduler instance.
Do not allocate the scheduler control structures from the workspace.
This is a preparation step for configuration of clustered/partitioned
schedulers on SMP.
This simplifies the RTEMS initialization and helps to avoid a memory
overhead. The workspace demands of the IO manager were not included in
the <rtems/confdefs.h> workspace size estimate. This is also fixed as a
side-effect.
Update documentation and move "Specifying Application Defined Device
Driver Table" to the section end. This sub-section is not that
important for the user. Mentioning this at the beginning may lead to
confusion.
Per task variables are inherently unsafe in SMP systems. This
patch disables them from the build and adds warnings in the
appropriate documentation and configuration sections.
Rename rtems_internal_error_description() to
rtems_internal_error_text(). Rename rtems_fatal_source_description() to
rtems_fatal_source_text(). Rename rtems_status_code_description() to
rtems_status_text(). Remove previous implementation of
rtems_status_text().
This patch includes a heavy rewrite of the chapter to have a more
structured approach to this chapter. It also changes the sectioning
to have the Data Structures be a section to themselves as a peer in the
outline with each logical area of macros in confdefs.h.
The 16 bit Object Id image was incorrectly rotated in the manual.
Both the 16 and 32 Object Id bit images had thin borders on the table
and the bottom line was not visible in the PDF. I tinkered with the
original drawings to add a bit more space around the tables and
manually converted the PNG to EPS using Gimp to ensure a controlled
and high quality conversion.
Do not return from rtems_initialize_start_multitasking() and call
rtems_fatal() instead with a fatal source of RTEMS_FATAL_SOURCE_EXIT and
a fatal code with the exit status.
Remove all bsp_cleanup() functions. The boot_card() is now a no return
function.
Add CPU port type CPU_Exception_frame and function
_CPU_Exception_frame_print().
The CPU ports of avr, bfin, h8300, lm32, m32c, m32r, m68k, nios2, sh,
sparc64, and v850 use an empty default implementation of
_CPU_Exception_frame_print().
Add rtems_exception_frame and rtems_exception_frame_print().
Add RTEMS_FATAL_SOURCE_EXCEPTION for CPU exceptions. Use rtems_fatal()
with source RTEMS_FATAL_SOURCE_EXCEPTION in CPU ports of i386, powerpc,
and sparc for unexpected exceptions.
Add third parameter to RTEMS_BSP_CLEANUP_OPTIONS() which controls the
BSP_PRINT_EXCEPTION_CONTEXT define used in the default
bsp_fatal_extension().
Add test sptests/spfatal26.
Call the fatal handlers of the user extensions before the update of
_Internal_errors_What_happened. This reduces the requirements on the
execution context further. Now a valid read-write data is only required
after the call to the fatal handlers.
