Bookmark and Share
Blog powered by Typepad
Member since 03/2007

Become a Fan

« DO-178B Quiz #2 | Main | Answers to DO-178B Quiz #2 »

February 02, 2010

Comments

Feed You can follow this conversation by subscribing to the comment feed for this post.

Outsourcing Press Kit Template

Thanks to give me these type of information thanks one again

property management software

I have just pondered the buckets of points you were trying to emphasize and I want to agree with you that there are a lot of things to not just be aware of but be scared and take a careful note. Thanks for the great tips you have just listed.

Association Management Software

When handling software one must always be careful so that there will be no problems. I tried once to use a particular software that I am not that familiar with and the result is that my computer got a worm that destroyed my important files.

Nicolas

Denis> see CAST-17 and CAST-12 document

What about the traitement of optimisation ? Some official FAQ say it's a different point from compiler-added code, because if the coverage does not change, no analyse of correctness is needed. But they wrote this with the short-circuit boolean evalution done in C in mind, which contain jump.

Optimisation is also a good way to have short function, that will help debugging and review. But if we can't count on inlinling, performance will suffer.

How justify inlining code and constant propagation ?

Denis

Amitabh, thanks a lot for your response. It made clear some aspects of the problem.

As I understood there is no "silver bullet" and there is no exact instructions about how to verify the compiler-introduced code. It depends on various conditions. OK, will investigate next ...

Denis

Amitabh

@Denis, How I wish I could give (or sell ;-)) some training material!

Anyways, to respond to your questions ... your manual analysis should reveal whether the compiler-introduced code has any impact on the structural coverage. It is just a additional statement that would have anyway got covered had you carried out the low-level testing on the object code instead of source code (with no additional input data). If yes, then you do not have to worry. However, if the object code is some sort of jump statement, then you have to verify the object code by giving break points, etc. You could also manually analyze the flow of the data. But the question you need to then ask is why did the compiler generate this code despite setting it to no-optimization. Or perhaps code needs to be rewritten to avoid that particular construct.

Hope this helps!

Denis

"Carry out a manual source code to assembly code comparison to ensure that the compiler has not introduced code (or deleted them) that cannot be traced back to the source code. (And no, you do not have to do this for the entire 50,000 lines of the software)"

Do you have some exact instructions or some training materials for source code - object code traceability investigation?

And what should I do when I found some object code that is introduced by compiler and that is not traced on the source code? DO-178B says about some "verification of the introduced object code" but what does it mean?

Thanks a lot

Denis

The comments to this entry are closed.

My Photo

Life sometimes turns on a dime. Especially when you are actively looking out for a change. I now work at HCL Technologies. And all I am looking for is to make a difference. By the way, the blog contents remain my own. It does not reflect the official position of HCL.

Your email address:


Powered by FeedBlitz