This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Import legacy CCS3.3 to CCS5.1 did not handle files

While trying to import CC3.3 project there were numerous errors that are not clear how to fix.

The messages in the project.log listed problems with the case of file names, here is an example:

!NOTE: Source file 'source\acu_28x_ADC.c' exists in the project, but with a different case. File is renamed to match the name in the PJT file.

So it would seem it fixed the problems but yet it then reported them as unresolved source files.  These files appear greyed out in CCS5 but double clicking on them show that it could still be opened.  I tried deleting one of these then adding to the project again but it did not help.  Attached is a screen shot.  (Note there is a \ vs /(backslash vs forward slash) difference but not sure why)

Any suggestions would be appreciated.

  • Hi Dan,
    There issue with older CCSv5 version in regards to poor handling of case in paths. This was resolved in later versions. Your issue may be related. In addition to that issue, later versions of CCS also has other various fixes to improve the legacy import wizard. Hence I strongly recommend moving to the latest version of CCS (v6.1.2). If you need to stay on CCSv5 for whatever reason, at least move to CCSv5.5.

    Thanks
    ki
  • Hi Ki,

    I installed CCS v6.1 and the legacy project import went much better, but there are still a few problems.

    There are a lot of *.IF files that it has trouble getting, I'm not should what these are.  Also the old compiler flags -b, -js, and -w were not translated into something equivalent.

    Below is a screen shot.

    Any suggestions would be great!

    Thanks,

    Dan

  • Hi Dan,
    The version of the compiler that comes with CCSv6 is much later than what was in CCSv3. Same with DSP/BIOS. Some options available with the old compiler are likely deprecated or unsupported with the newer compiler.

    You have two options:
    1) Use the same version of the compiler used with CCSv3.3. You can choose to download/install a version of the same compiler version used with CCSv3.3 and point CCSv6 to use it (or if you already have CCSv3.3 installed on the same PC, you can just use the compiler in your CCSv3.3 installation). Directions on how to do this is mentioned in:
    processors.wiki.ti.com/.../Compiler_Installation_and_Selection

    2) You can try tweaking the project settings to set the options of the new compiler to work similar to the options of the old compiler.

    I'd go with #1 for now to make the migration easier.

    I also see that you have some warnings with DSP/BIOS also. What version were you using with v3.3? There shouldn't be much difference between BIOS 5.3x and 5.4x in regards to building it. In any case, it looks like your build failed on a source file compiler error so see if switching to the original compiler version helps.

    Thanks
    ki
  • Hi Ki,

    You asked which DSP/BIOS version I was using with v3.3, but I was not using 3.3.  When I joined the company I was given a Windows 7 PC and they told me that CCS3.3 will not run on WIndows 7, so I installed CCS5.5 and started a new project from scratch.  Now I need to work on an old project that the persons who developed it are no longer with the company.

    I prefer not to go back to the old 3.3 compiler if possible.  This should be fixable.

    It seems that there is really just one compiler error with the *.IF files(which is repeated many times)

    Is the "IF" short for "intermediate file"?  If so can you tell me what step or tool would build this file?

    Maybe the intermediate file is not built or in the correct directory.  It might be a simple fix.  If I could get the entire compile/build process to stop when it has this error, it might be helpful.  It might be a path problem or environment variable.

    Once this is solved then I can go forward.  The warnings are less important.

    I just searched and saw that other people who had the infamous "Fatal Error #1965:" had a lot of trouble trying to fix it(that's scary!)

    I am looking thru the makefile to try to follow the flow.

    Here is another more readable screenshot...