Skip to contentSkip to conversation

LTP scheduler testing

LTP Subrata Modak 1 participant 1 message Jun 23, 2009 2 messages not in the archive

lore
Share & tools
lore

A message that is not in the archive (sent off-list or not yet received): [email protected]

A message that is not in the archive (sent off-list or not yet received): [email protected]

#2 lore raw
Quoted 9 lines · Subrata Modak

Hi Mac,

Quoted 6 lines · Mac Mollison

Hi,

I'm working with a group researching real-time systems at the University of North Carolina at Chapel Hill. I'm writing a testing framework for a linux kernel scheduler patch called LITMUS^RT, which allows us to test out various scheduling policies that are not in the normal kernel. I'm

Are you still working on this ? And, can you share the tests with us ?

Regards--
Subrata

Quoted 54 lines

Good to know about your work.

Quoted 4 lines

also writing a paper about my efforts. I have a few questions...

1) Have any papers been published discussing LTP's testing of the Linux kernel scheduler (including the real-time parts)?

Nope. Papers have been published about the general testing (including schedulers) through LTP, but, nothing specific to Schedulers alone.

Quoted 3 lines

2) As far as I can tell, the LTP is not concerned with testing that some scheduling policy is followed, but just that the scheduler is 'working'.

True. We are concerned only about that the scheduler that your kernel is built/enabled/booted with is working fine. And you can verify whether any subsequent kernel commits has introduced any significant regression.

Quoted 5 lines

For example, you wouldn't be trying to ensure that the completely fair scheduler is actually completely fair, just that it appears to be scheduling things properly. Is this correct? (My research is more interested in whether or not a theoretical scheduling policy is being followed in practice, not whether or not it "works" for the end user.)

I would be also interested in getting your work tested by the LTP community. We would be really interested in something like:

1) Determine which scheduling policy is employed in the running kernel (through /boot/config* or /pros/config*),
2) Then invoke the corresponding testcase to find whether the job is being done correctly.

Quoted 3 lines

3) Are you aware of other projects which have done work on testing scheduling in linux, or any published papers on the topic?

We have the REALTIME tests in LTP, which measures all those aspects, but only for the RT kernel. You can check them out here. Hope this helps:

http://ltp.cvs.sourceforge.net/viewvc/ltp/ltp/testcases/realtime,

Regards--
Subrata

Quoted 7 lines

Thanks so much for your time and help!

Mac Mollison

Quoted 5 lines

------------------------------------------------------------------------------
_______________________________________________
Ltp-list mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ltp-list

------------------------------------------------------------------------------
Are you an open source citizen? Join us for the Open Source Bridge conference! Portland, OR, June 17-19. Two days of sessions, one day of unconference: $250. Need another reason to go? 24-hour hacker lounge. Register today!
http://ad.doubleclick.net/clk;215844324;13503038;v?http://opensourcebridge.org
_______________________________________________
Ltp-list mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ltp-list