litmus2008.git/kernel, branch RTSS08 [ARCHIVE] Old LITMUS^RT 2008 version (for reference). Feather-Trace: make SCHED only depend on next 2008-05-22T16:41:46+00:00 Bjoern B. Brandenburg bbb@cs.unc.edu 2008-05-22T16:41:46+00:00 660e3e3773f63c99c4eaabed3a49344628719e29 






record task type in feather trace time stamp
2008-05-22T15:49:47+00:00

Bjoern B. Brandenburg
bbb@cs.unc.edu

2008-05-22T15:49:47+00:00

4e94db6f1c0341c5e64f0f1487de63420fb7a163












Add Feather-Trace support.
2008-05-22T04:47:29+00:00

Bjoern B. Brandenburg
bbb@cs.unc.edu

2008-05-22T03:01:47+00:00

ee869ea3b9715e81d75288760011080e82244701

- TICK
- SCHED1
- SCHED2
- RELEASE
- CXS





- TICK
- SCHED1
- SCHED2
- RELEASE
- CXS







TRACE(): facilitate debugging
2008-05-11T22:28:09+00:00

Bjoern B. Brandenburg
bbb@cs.unc.edu

2008-05-11T22:28:09+00:00

1b0b2b02697583bbe3aa6060e3f28d0a75c66ab6

- don't deadlock on a runqueue lock in case of a bug inside
  the scheduler
- write printk() messages to TRACE()
- tell user of TRACE() buffer





- don't deadlock on a runqueue lock in case of a bug inside
  the scheduler
- write printk() messages to TRACE()
- tell user of TRACE() buffer







two fixes and some logging improvement
2008-05-10T08:31:43+00:00

Bjoern B. Brandenburg
bbb@cs.unc.edu

2008-05-10T08:31:43+00:00

1939631637ed7ac662126a76ef7665e8295ac61d

- remove outdated comment
- reorder stack_in_use marker to be in front of finish_lock_switch()
- add TRACE()s to try_to_wake_up()





- remove outdated comment
- reorder stack_in_use marker to be in front of finish_lock_switch()
- add TRACE()s to try_to_wake_up()







LITMUS: disable Linux scheduler optimization
2008-05-05T18:27:26+00:00

Bjoern B. Brandenburg
bbb@cs.unc.edu

2008-05-05T18:27:26+00:00

8f5fecd0290df1bba43f17260534261f6689a39e

This caused LITMUS^RT real-time tasks to be missed, which
in turn caused all kinds of inconsistent state.





This caused LITMUS^RT real-time tasks to be missed, which
in turn caused all kinds of inconsistent state.







LITMUS: avoid using the same stack on two CPUs in global schedulers
2008-05-04T22:07:35+00:00

Bjoern B. Brandenburg
bbb@cs.unc.edu

2008-05-04T22:07:35+00:00

6b06d1ce233787655eb21b624ed924806768b36c

This change fixes a race where a job could be executed on more than one
CPU, which to random crashes.





This change fixes a race where a job could be executed on more than one
CPU, which to random crashes.







LITMUS: don't let real-time tasks fork real-time tasks.
2008-05-03T17:22:36+00:00

Bjoern B. Brandenburg
bbb@cs.unc.edu

2008-05-03T17:22:36+00:00

9b4ccbc27ea5d1a35e79391ca5a500b32cd253a1

The forked task will be a best-effort task.





The forked task will be a best-effort task.







LITMUS: add framework for carrying out jobs after dropping rq locks
2008-05-03T02:53:02+00:00

Bjoern B. Brandenburg
bbb@cs.unc.edu

2008-05-03T02:53:02+00:00

ae9203a19764fd434ded1e4b16ce897566aba63c

Many things can't be done from within the scheduler.
This framework should make it easer to defer them.





Many things can't be done from within the scheduler.
This framework should make it easer to defer them.







Merge commit 'jupiter/debug' into debug
2008-05-03T00:52:04+00:00

Bjoern B. Brandenburg
bbb@cs.unc.edu

2008-05-03T00:52:04+00:00

d8be518053ec5b66250836f6989d1dffca5b021c