padmanabham:
It is going to be rather hard to continue the discussion on the use of hard coded subroutine addresses unless you can share with us why you feel that this is a necessary thing to do!!

I wrote before, and will say again, that it is really unnecesary to hard code subroutine addresses. On top of that it is generally not the best practice from the standpoint of code support and maintenance.

You indicate that you have gone ahead and placed one or more C-functions at specific addresses. The literal fact that you do this leads exactly to the warning you specify that you may encounter conflicts of the code memory usage. Think ahead to 7 or 8 years from now where some new person has the job of making changes to the code in this product. It may be that you will not even be there to tell him about why you hard coded the subroutine addresses. If he increases the code size he may encounter problems and have very much difficulty.

I would suggest that the best long term solution is to setup the build process for the project so that it is possible to use a very simple procedure with the very most minimal "special considerations". This way the future code support and maintenance is easy. They just add/change code and then follow some simple instructions to compile / assemble / link and download or program new parts. It is best if this process is done off a single folder set of source files and that the whole code for the project be built in one build step!! Hard coded function subroutine addresses often imply that one of the above recommendations is not being followed.

Michael Karas