Add priority nodes which contribute to the overall thread priority.
The actual priority of a thread is now an aggregation of priority nodes.
The thread priority aggregation for the home scheduler instance of a
thread consists of at least one priority node, which is normally the
real priority of the thread. The locking protocols (e.g. priority
ceiling and priority inheritance), rate-monotonic period objects and the
POSIX sporadic server add, change and remove priority nodes.
A thread changes its priority now immediately, e.g. priority changes are
not deferred until the thread releases its last resource.
Replace the _Thread_Change_priority() function with
* _Thread_Priority_perform_actions(),
* _Thread_Priority_add(),
* _Thread_Priority_remove(),
* _Thread_Priority_change(), and
* _Thread_Priority_update().
Update #2412.
Update #2556.
This patches some issues with the capture engine:
1. Check is the engine is open in ctrace commands.
2. Check all record open and appends for overflow.
3. Fix the record open to take the size of user data and
not the record header.
4. Use packed structs for data being written to the per
cpu buffers.
5. Remove direct struct access to the capture buffers to
avoid misaligned accesses.
6. Add support to extract records, no struct access to the
capture buffers.
7. Update ctrace to extract records from the capture buffers.
8. Add support to ctrace to always print the task name if it
has one.
9. Add support to manage names or the lack of a name.
10. Range of minor fixes.
11. Fix a long standing bug in ctset's handling of args.
Closes#2780.
The printing of 64bit diff's for the timestamp did not work on ARM
so I have changed this to 32bit. A 32bit nano-second diff between
events should be more than enough time.
Print '____' for POSIX thread output.
Methods to print the data were moved from capture-cli into
a support area and are no longer static so that they can
be shared by test routines, or application code that wants
to use the capture engine without the shell interface.
The catpture task record is now logged just prior to the first
log entry using that task instead of the first time the task
is seen. This involved splitting the record task method into
an initialize task and a record task.
This patch removes functionality for stack checking from
the capture engine and requiresi the use of existing rtems
functions for this information. It modifies ctload to use
functionality similar to rtems cpuusage. It removes the
capture task and stores a new capture task record the first
time the task is seen. The per task data that was still
needed is scaled down and stored in the tcb.
The function to change a thread priority was too complex. Simplify it
with a new scheduler operation. This increases the average case
performance due to the simplified logic. The interrupt disabled
critical section is a bit prolonged since now the extract, update and
enqueue steps are executed atomically. This should however not impact
the worst-case interrupt latency since at least for the Deterministic
Priority Scheduler this sequence can be carried out with a wee bit of
instructions and no loops.
Add _Scheduler_Change_priority() to replace the sequence of
- _Thread_Set_transient(),
- _Scheduler_Extract(),
- _Scheduler_Enqueue(), and
- _Scheduler_Enqueue_first().
Delete STATES_TRANSIENT, _States_Is_transient() and
_Thread_Set_transient() since this state is now superfluous.
With this change it is possible to get rid of the
SCHEDULER_SMP_NODE_IN_THE_AIR state. This considerably simplifies the
implementation of the new SMP locking protocols.
Print the uptime and difference in nanoseconds to the previous
record in the ctrace output. For example:
0:00:59.474927121 14760 0a01000c TNTD 235 235 CREATED_BY
0:00:59.474927418 297 0a010012 /dev 235 235 CREATED
0:00:59.474930799 3381 0a01000c TNTD 235 235 STARTED_BY
0:00:59.474931105 306 0a010012 /dev 235 235 STARTED
0:00:59.475072297 141192 0a01000c TNTD 235 235 SWITCHED_OUT
The capture did not work due to changes in the workspace allocator.
The engine now scans all existing tasks when enabled and does any
allocations then.
Fixed a bug in the ctset commandi in the CLI.
Updated the capture engine to use 64bit nanosec timestamps.
Fixed the CLI showing the stack usage.
Move implementation specific parts of states.h and states.inl into new
header file statesimpl.h. The states.h contains now only the
application visible API.
Move implementation specific parts of watchdog.h and watchdog.inl into
new header file watchdogimpl.h. The watchdog.h contains now only the
application visible API.
Script does what is expected and tries to do it as
smartly as possible.
+ remove occurrences of two blank comment lines
next to each other after Id string line removed.
+ remove entire comment blocks which only exited to
contain CVS Ids
+ If the processing left a blank line at the top of
a file, it was removed.
* libmisc/capture/capture.c, posix/src/keyfreememory.c,
posix/src/pthread.c, score/include/rtems/score/wkspace.h,
score/src/objectextendinformation.c,
score/src/objectnamespaceremove.c, score/src/objectsetname.c,
score/src/threadclose.c, score/src/threadinitialize.c,
score/src/wkspace.c: Many places were checking for a NULL pointer
before calling _Workspace_Free. By moving the check into
_Workspace_Free, we eliminate a number of conditional paths and make
it harder to return a NULL pointer.
