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.

Non-repeatable Compile Results



Still Having Fun?

 

I'm getting non-repeatable compiler results..  Geting the error:


--retain=g_pfnVectors not supported when producing TI-COFF output files        ez_Host    line 0    1281312357609    1913

I see that  --retain=g_pfnVectors is in the 'cmd' file but even commented out it produces the error

Default configure processes an ERROR.. 

 

 

  • David,

    Could you provide more details on the errors? I assume you are changing the output format from EABI(ELF) to legacy COFF and commenting out the --retain line from the linker command file. Are you saying you get different results for repeated builds with these changes?

  • The --retain option is valid only when building for EABI (ELF).

    Thanks and regards,

    -George

     

  • I didn't change the output, can't I'm running a x64 OS and those options never appear, the IDE crashed and when I restarted it and compiled I got errors. After spending hours playing with the .cdtbuild file including 'setting build defaults' I got it to compile OK. Setting build defaults generated the same error also.  The '--retain...' line in .cdtbuild was commented out. I got successful compile when I used copied the .cdtbuild from a common/similar project.  

    The fun got better then...  I started to see the FaultISR getting tripped during safeRTOS calls, but after some removing of non safeRTOS code I was able to get a compiled piece of code to run, BUT not always. To make things even better the same project compiles and runs smooth when compiled with Keil compiler (project taken from EKK-LM3S9B96).

    I have also have seen not  been able to set breakpoints, received AET error when attempting to load, crashed. And one that is thr most troubled is that when loading the program I will see the 'Loading' window about 15 seconds after I activate the operation. And then it will stay at 2% complete for about 30 seconds and then jump to about 80% then it's completed (so it says)

    May last adventure I recompiled the some code that ran before but it now fails every time (FaultIRS)

  • George

     

    The compile configuration was set to default after a CCS crash. Also the --retain line was commented out in the .cdtbuild  file.  I can't see any build configuration data since CCS does not run successful on the my x64 bit OS. I can only peek into the .cdtbuild file but it appears that this is not the only file that drives the build.

     

     

    I guess everything works great when there is not any problems.

  • This whole thread is confusing to me.  It ranges across several topics.  Not all of them are compiler topics I can help with.  I'll lend what insight I can.

    As to the original post which gives the linker error message about combining --retain and TI-COFF ...  It seems you are building a Stellarisware project.  When I look at my Stellarisware installation (which is not the latest version), all the projects use EABI and none of them use --retain.  So, I don't understand how your Stellarisware project is building for TI-COFF and using --retain.

    I don't know anything about the FaultISR or safeRTOS calls.  Though I did find this http://www.highintegritysystems.com/index.php?option=com_content&view=article&id=93&Itemid=97 .

    For more information on how CCS does builds, or the AET questions, I recommend the CCS forum: http://e2e.ti.com/support/development_tools/code_composer_studio/f/81.aspx .

    Sorry I couldn't be more help.

    -George

     

  • This was fixed once I was able to see the Build options