Home > Archive > Software Engineering > May 2006 > MDA and chronometric Analysis
You are viewing an archived Text-only version of the thread.
To view this thread in it's original format and/or if you want to reply to
this thread please [click here]
| Author |
MDA and chronometric Analysis
|
|
| David Lightstone 2006-05-07, 10:05 pm |
| Seems to me that you can't do Hard real-time reliably unless you can do a
chronometric analysis on the developed code. (i.e. determine the various
execution thread segments and their worst case execution times, followed by
setting appropriate priorities for the scheduler of choice).
Do any of the MDA tools support this activity?
| |
| S Perryman 2006-05-07, 10:05 pm |
| "David Lightstone" <david._NoSpamlightstone@prodigy.net> wrote in message
news:JL%6g.6035$mu2.531@newssvr24.news.prodigy.net...
> Seems to me that you can't do Hard real-time reliably unless you can do a
> chronometric analysis on the developed code. (i.e. determine the various
> execution thread segments and their worst case execution times, followed
> by setting appropriate priorities for the scheduler of choice).
> Do any of the MDA tools support this activity?
In 2002, I think it was the makers of the Artisan CASE tool who stated that
they had integrated (or were integrating) an RMA tool into their tool chain.
Given the elapsed timespan, I assume that you should be able to find a set
of
vendors on the MDA trip (Rhapsody etc) who have some story for hard-real
time schedulability etc.
Hope this helps.
Regards,
Steven Perryman
| |
| H. S. Lahman 2006-05-07, 10:05 pm |
| Responding to Lightstone...
> Seems to me that you can't do Hard real-time reliably unless you can do a
> chronometric analysis on the developed code. (i.e. determine the various
> execution thread segments and their worst case execution times, followed by
> setting appropriate priorities for the scheduler of choice).
>
> Do any of the MDA tools support this activity?
It depends on how you define 'support'. B-)
The short answer is: No. Typically threads and hard R-T constraints are
defined in an MDA Marking Model for the application. That is usually
done by tagging model elements in the drawing tools with "pass-through"
tags. The transformation engine can then do the grunt work for things
like invoking OS threading facilities, optimizing blocking, implementing
a scheduler, etc.
A "smart" tool will minimize the information the developer needs to
supply in the Marking Model. However, the developer ultimately needs to
perform the basic analysis that you describe.
FYI, all of the classic translation tools -- Bridgepoint, PathMATE,
Rhapsody, Rose/RT, xUML -- provide code generation facilities of some
sort for hard R-T because all those tools originally grew up in R-T/E.
*************
There is nothing wrong with me that could
not be cured by a capful of Drano.
H. S. Lahman
hsl@pathfindermda.com
Pathfinder Solutions -- Put MDA to Work
http://www.pathfindermda.com
blog: http://pathfinderpeople.blogs.com/hslahman
Pathfinder is hiring:
http://www.pathfindermda.com/about_us/careers_pos3.php.
(888)OOA-PATH
| |
| David Lightstone 2006-05-07, 10:05 pm |
|
"H. S. Lahman" <h.lahman@verizon.net> wrote in message
news:t5q7g.766$052.760@trnddc02...
> Responding to Lightstone...
>
>
> It depends on how you define 'support'. B-)
Right now I would be happy if it could simply enumerate all path segments
each individual path segment being defined by 2 sequencial context switches
>
> The short answer is: No. Typically threads and hard R-T constraints are
> defined in an MDA Marking Model for the application. That is usually done
> by tagging model elements in the drawing tools with "pass-through" tags.
> The transformation engine can then do the grunt work for things like
> invoking OS threading facilities, optimizing blocking, implementing a
> scheduler, etc.
>
> A "smart" tool will minimize the information the developer needs to supply
> in the Marking Model. However, the developer ultimately needs to perform
> the basic analysis that you describe.
>
> FYI, all of the classic translation tools -- Bridgepoint, PathMATE,
> Rhapsody, Rose/RT, xUML -- provide code generation facilities of some sort
> for hard R-T because all those tools originally grew up in R-T/E.
>
> *************
> There is nothing wrong with me that could
> not be cured by a capful of Drano.
>
> H. S. Lahman
> hsl@pathfindermda.com
> Pathfinder Solutions -- Put MDA to Work
> http://www.pathfindermda.com
> blog: http://pathfinderpeople.blogs.com/hslahman
> Pathfinder is hiring:
> http://www.pathfindermda.com/about_us/careers_pos3.php.
> (888)OOA-PATH
>
>
>
| |
| H. S. Lahman 2006-05-10, 7:03 pm |
| Responding to Lightstone...
>
>
>
> Right now I would be happy if it could simply enumerate all path segments
> each individual path segment being defined by 2 sequencial context switches
Don't you get that part "for free" in a UML Interaction Diagram? Each
message triggers a context switch at the method invocation level. A
sequence Diagram provides sequencing in time while a Collaboration
Diagram provides logical sequencing.
Typically people "colorize" threads on the Interaction Diagram. (The
term comes from the Olden Days when one used different colors of
highliters on hardcopy to identify threads.) The tricky part is
deciding how many segments to include. B-)
*************
There is nothing wrong with me that could
not be cured by a capful of Drano.
H. S. Lahman
hsl@pathfindermda.com
Pathfinder Solutions -- Put MDA to Work
http://www.pathfindermda.com
blog: http://pathfinderpeople.blogs.com/hslahman
Pathfinder is hiring:
http://www.pathfindermda.com/about_us/careers_pos3.php.
(888)OOA-PATH
| |
| David Lightstone 2006-05-10, 7:03 pm |
|
"H. S. Lahman" <h.lahman@verizon.net> wrote in message
news:pZq8g.8783$k%2.2471@trnddc03...
> Responding to Lightstone...
>
>
> Don't you get that part "for free" in a UML Interaction Diagram? Each
> message triggers a context switch at the method invocation level. A
> sequence Diagram provides sequencing in time while a Collaboration Diagram
> provides logical sequencing.
I can not give a definitive answer. If the Intreraction Diagram is at the
level of tasks interacting, definitely.
I can not conclusively state however that the interaction diagram will be
prepared at the task level
>
> Typically people "colorize" threads on the Interaction Diagram. (The term
> comes from the Olden Days when one used different colors of highliters on
> hardcopy to identify threads.) The tricky part is deciding how many
> segments to include. B-)
>
>
> *************
> There is nothing wrong with me that could
> not be cured by a capful of Drano.
>
> H. S. Lahman
> hsl@pathfindermda.com
> Pathfinder Solutions -- Put MDA to Work
> http://www.pathfindermda.com
> blog: http://pathfinderpeople.blogs.com/hslahman
> Pathfinder is hiring:
> http://www.pathfindermda.com/about_us/careers_pos3.php.
> (888)OOA-PATH
>
>
>
>
|
|
|
|
|