Use mostly the standard watchdog operations. Use a system event for
synchronization. This implementation is simpler and offers better SMP
performance.
Close#2131.
Rewrite the _Watchdog_Insert(), _Watchdog_Remove() and
_Watchdog_Tickle() functions to use iterator items to synchronize
concurrent operations. This makes it possible to get rid of the global
variables _Watchdog_Sync_level and _Watchdog_Sync_count which are a
blocking point for scalable SMP solutions.
Update #2307.
Add watchdog header parameter to _Watchdog_Remove() to be in line with
the other operations. Add _Watchdog_Remove_ticks() and
_Watchdog_Remove_seconds() for convenience.
Update #2307.
Avoid the usage of the current thread state in
_Thread_queue_Extract_with_return_code() since thread queues should not
know anything about thread states.
Account for priority changes of threads executing in a foreign
partition. Exchange idle threads in case a victim node uses an idle
thread and the new scheduled node needs an idle thread.
The cpuuse top command now supports the current load where the list of
tasks is ordered based on the current load rather than the total cpu usage.
This lets you see what is using the processor at any specific instance.
The ability to sort on a range of thread values is now supported.
Added memory usage stats for unified and separate workspace and C heaps as
well as displaying the allocated stack space.
Added a few more command keys to refresh the display, show all tasks in the
system, control the lines display and a scrolling mode that does not clear
the display on each refresh.
Removed support for tick kernel builds. The tick support in the kernel is to
be removed.
A thread join is twofold. There is one thread that exists and an
arbitrary number of threads that wait for the thread exit (one-to-many
relation). The exiting thread may want to wait for a thread that wants
to join its exit (STATES_WAITING_FOR_JOIN_AT_EXIT in
_POSIX_Thread_Exit()). On the other side we need a thread queue for all
the threads that wait for the exit of one particular thread
(STATES_WAITING_FOR_JOIN in pthread_join()).
Update #2035.
There may be a way to reduce the memory requirements but it
will require time to ensure the math is right and it passes
on all targets. At the current time, it fails on 22 BSPs which
run on simulators.
Move the linear search into a critical section to avoid corruption due
to higher priority interrupts. The interrupt disable time depends now
on the count of pending messages.
Close#2328.
