I updated the benchmark to include also the RTEMS chain implementation
for comparison with the red-black trees. Below 60 nodes it is faster
than the red-black trees.
https://github.com/sebhub/rb-bench
Is a red-black tree really the best choice for our priority queues? A
semaphore with 60 waiting tasks is likely a problem itself in an
application.
On 17/12/14 10:46, Sebastian Huber wrote:
Hello,
I added some very simple benchmark code here:
https://github.com/sebhub/rb-bench
On 06/11/14 07:53, Sebastian Huber wrote:
Hello,
I have a new item for the list:
Very desirable
==============
+ Since red-black trees are now used to implement the priority queues
and they will play an important part in future SMP improvements I
would like to do some performance measurements with alternative
implementations. I would like to compare
1. the RTEMS red-black tree implementation,
2. the RTEMS red-black tree implementation with the colour encoded in
the parent pointer,
3. the implementation from Eternally Confuzzled, see
http://www.eternallyconfuzzled.com/tuts/datastructures/jsw_tut_rbtree.aspx,
4. the BSD implementation <sys/tree.h>, and
5. the JFFS2 implementation
(cpukit/libfs/src/jffs2/include/linux/rbtree.h), see
http://lwn.net/Articles/184495/.
I intend to use real hardware ARM, ColdFire, PowerPC and SPARC targets.
Do we have other interesting red-black tree implementations?
On 29/10/14 22:49, Joel Sherrill wrote:
Hi
Since we are long over due for a release. I can count 4 new children
and a grandchild among the core developers since 4.10. :)
Here is what I have off the top of my head:
Mandatory
==========
+ Tool version selection
- Likely gcc 4.9 on almost all targets.
- binutils last release
- newlib TBD
- GDB could be new release
+ BeagleBoard BSP added
+ Run-Time Loader merged and tested
+ Visit the Getting Started manual. It is dated.
+ Push a bit more on warnings.
- Three BSPs represent bulk of remaining warnings
Very desirable
=============
+ x86 context switch SMP handoff logic
+ Pi GSOC Code merged.
Anything else?
--
Sebastian Huber, embedded brains GmbH
Address : Dornierstr. 4, D-82178 Puchheim, Germany
Phone : +49 89 189 47 41-16
Fax : +49 89 189 47 41-09
E-Mail : sebastian.hu...@embedded-brains.de
PGP : Public key available on request.
Diese Nachricht ist keine geschäftliche Mitteilung im Sinne des EHUG.
_______________________________________________
devel mailing list
devel@rtems.org
http://lists.rtems.org/mailman/listinfo/devel
