[Bug 1647] Modular SuperCore Scheduler

bugzilla-daemon at rtems.org bugzilla-daemon at rtems.org
Sat Nov 13 18:27:20 UTC 2010


https://www.rtems.org/bugzilla/show_bug.cgi?id=1647

--- Comment #50 from Gedare <giddyup44 at yahoo.com> 2010-11-13 12:27:19 CST ---
The new patch and tarball are now up.  I have also posted results of tmtests
and sptests for both sis and psim.  I fixed the sizing issue with confdefs.h

The only substantial time increases that I see are in the task create and task
delete timings.  As I indicated before, I'm pretty sure this is due to new
workspace allocations.  It amounts to about 100 instructions according to psim.

The general scheduling interface is that only the functions and structures
declared in rtems/score/scheduler.h should be accessed from outside of the
scheduler. There remain two places in the score Thread management code that are
not properly encapsulated: threadsettransient.c and threadchangepriority.c both
make direct modifications of the priority scheduler ready queue.  Down the line
it may become necessary to implement a Scheduler Handler interface to the ready
queue for these purposes, or at least for the thread extraction of the
Set_transient function.

Joel: The code has changed a bit, I collapsed the readyq code entirely into the
scheduler, so that schedulerpriority.inl now contains all of the ready queue
management which used to be in various readyqpriority.* files.  This may have
introduced style or code documentation regressions, but nothing that affects
the code functionality.

-- 
Configure bugmail: https://www.rtems.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are watching all bug changes.



More information about the bugs mailing list