Forums

Process vs Task model benefits for RTOS-mainly vxworks!-Let me understand this plz!

Started by ssubbarayan August 10, 2005
Gurus,
Till vxworks5.5,windriver guys have been using task based model in flat
address mode for application scheduling,but now with vxworks 6.0
windriver is playing with idea of using process model over tasks.
My question is what could be rational behind this?AFAIK,there were not
much of problems with task models and only advantage of going for
process model is enabling memory protection.But the adverse effect of
this would be timing impact due to MMU calculations every time context
switch happens and Windriver accepts there is performance hit in terms
of timing but still its deterministic.I am not able to understand how
this would be deterministic?And how far we can benefit from memory
protected applications?Because prior versions of vxworks were able to
achieve good performance even with out memory protection.

Guys using Linux whats your opinion on this?

Can some one explain me this?

Regards,
s.subbarayan

Hello,

There were some great white papers written up to explain Wind Rivers
memory and process model in 6.0.

http://www.windriver.com/portal/server.pt/gateway/PTARGS_0_39823_381175_0_0_18/rtps_for_vxworks6_wp.pdf

http://www.windriver.com/portal/server.pt/gateway/PTARGS_0_39823_381177_0_0_18/vxworks_memory_allocation_wp.pdf

Your local Account Team can also provide you with benchmarks regarding
the MMU overhead w.r.t. context switching.  Realize that the kernel is
always mapped into your local context so it's not like you have to load
all the TLB entries when moving between RTPs...only the user level
pages have to be loaded/unloaded; the overhead I've see is minimal on
all PPC architecture especially when compared to the savings in
development / maintenance when problems occur.

Jamie

ssubbarayan wrote:
> Gurus, > Till vxworks5.5,windriver guys have been using task based model in flat > address mode for application scheduling,but now with vxworks 6.0 > windriver is playing with idea of using process model over tasks. > My question is what could be rational behind this?AFAIK,there were not > much of problems with task models and only advantage of going for > process model is enabling memory protection.But the adverse effect of > this would be timing impact due to MMU calculations every time context > switch happens and Windriver accepts there is performance hit in terms > of timing but still its deterministic.I am not able to understand how > this would be deterministic?And how far we can benefit from memory > protected applications?Because prior versions of vxworks were able to > achieve good performance even with out memory protection. > > Guys using Linux whats your opinion on this? > > Can some one explain me this? > > Regards, > s.subbarayan
Try the following short cuts, if they don't work than search off the
main www.windriver.com page for "mem white paper" and "rtp white paper"

http://www.windriver.com/portal/server.pt?space=Opener&control=OpenObject&cached=true&parentname=SearchResult&parentid=4&in_hi_ClassID=18&in_hi_userid=27106&in_hi_ObjectID=381175&in_hi_OpenerMode=2&

http://www.windriver.com/portal/server.pt?space=Opener&control=OpenObject&cached=true&parentname=SearchResult&parentid=3&in_hi_ClassID=18&in_hi_userid=27106&in_hi_ObjectID=381175&in_hi_OpenerMode=2&

On 10 Aug 2005 07:16:57 -0700, "Jamie" <jamiehelmer@hotmail.com>
wrote:

>http://www.windriver.com/portal/server.pt/gateway/PTARGS_0_39823_381175_0_0_18/rtps_for_vxworks6_wp.pdf > >http://www.windriver.com/portal/server.pt/gateway/PTARGS_0_39823_381177_0_0_18/vxworks_memory_allocation_wp.pdf
I get: Error - The server has experienced an error. Try again, or contact your Portal administrator if you continue experiencing problems. Problem? Jon
On 10 Aug 2005 08:39:39 -0700, "Jamie" <jamiehelmer@hotmail.com>
wrote:

>Try the following short cuts, if they don't work than search off the >main www.windriver.com page for "mem white paper" and "rtp white paper" > >http://www.windriver.com/portal/server.pt?space=Opener&control=OpenObject&cached=true&parentname=SearchResult&parentid=4&in_hi_ClassID=18&in_hi_userid=27106&in_hi_ObjectID=381175&in_hi_OpenerMode=2& > >http://www.windriver.com/portal/server.pt?space=Opener&control=OpenObject&cached=true&parentname=SearchResult&parentid=3&in_hi_ClassID=18&in_hi_userid=27106&in_hi_ObjectID=381175&in_hi_OpenerMode=2&
Those are both the same link. But they do work for one of the papers (not the other one.) Thanks. Jon
ssubbarayan wrote:
> Gurus, > Till vxworks5.5,windriver guys ...
<snip> You might be able to achieve an effect similar to 5.5 by running all your threads in _one_ process. After all, 5.5 is just one big anonymous process that contains all those 5.5 threads.
ssubbarayan wrote:
> Gurus, > Till vxworks5.5,windriver guys have been using task based model in flat > address mode for application scheduling,but now with vxworks 6.0 > windriver is playing with idea of using process model over tasks. > My question is what could be rational behind this?AFAIK,there were not > much of problems with task models and only advantage of going for > process model is enabling memory protection.But the adverse effect of > this would be timing impact due to MMU calculations every time context > switch happens and Windriver accepts there is performance hit in terms > of timing but still its deterministic.I am not able to understand how > this would be deterministic?And how far we can benefit from memory > protected applications?Because prior versions of vxworks were able to > achieve good performance even with out memory protection. > > Guys using Linux whats your opinion on this? > > Can some one explain me this? > > Regards, > s.subbarayan >
In my humble opinion, and still after reading the white papers, it is just a means for lazy or incapable programmers, which adds overhead. You can develop your code with the MMU on and after some testing and verifying, that there is no memory violation (program error) you can switch the MMU off and after the final testing ship your product without the unnecessary overhead. I wish I could do the same thing with Linux:) Regards, Robert --- Robert Berger Embedded Systems Software Group Leader INTRACOM S.A. Hellenic Telecommunications & Electronics Industry Content Delivery Systems Department P.O. Box 68, 19.5 km Markopoulo Ave. (Building A) 190 02 Peania, Athens, GREECE Tel.: (+30 210) 667 4353, 667 9000 Fax.:(+30 210) 667 7101 email: rber@intracom.gr URL: www.intracom.gr --- ...Under a government which imprisons any unjustly, the true place for a just man is also a prison. -- Henry David Thoreau My public pgp key is available at: http://pgp.mit.edu:11371/pks/lookup?op=get&search=0x90320BF1 The information contained in this e-mail and in any attachments is confidential and is designated solely for the attention of the intended recipient(s). If you are not an intended recipient, you must not use, disclose, copy, distribute or retain this e-mail or any part thereof. If you have received this e-mail in error, please notify the sender by return e-mail and delete all copies of this e-mail from your computer system(s).