• State Resolved
  • Locked Locked
  • Replies 7 replies
  • Subscribers 54 subscribers
  • Views 561 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

CCSv5 on linux

Michael Jordan
Intellectual 955 points

Other Parts Discussed in Thread: OMAPL138

My ccsv5.4.0.00091 installation was becoming quite unreliable. I finally got to the point of deleting it, deleting the ccsv5 directory and the InstallJammer directory as well. I cleaned out every bit of reference/configuration I had that looked like it could be connected. Then I reinstalled. The problems continued. Eventually I discovered something which may be of help to someone down the line. Working with many microcontrollers at school and work, I deal with a variety of IDE's. One of which was MPLabX by Microchip. They have a cool, slick netbeans based IDE. However, I got to looking around and noticed that my udev rules didn't seem to be implementing properly. Microchip (Microsoft, Microshaft, whatever), it turns out, was to blame. With their installation, they create a hidden directory in /etc. Then they place inside it an executable file. They then set up to take control of anything that might even look like a serial port! I listed my udev database, and every tty*, every vc*, every ttyUSB*, etc. were all showing as following the configuration in this hidden directory executable. Once I finally discovered the root of the problem, I uninstalled MPLabX and removed the hidden directory. After notifying udev of the changes and rebooting, I finally got all of my devices back. EVERYTHING on the system now appears to work smoother now. This includes ccsv5, once I had installed it again.

Having said that, I still have one annoying issue which I was hoping perhaps someone could help with.

When I debug with my EXPKITOMAPL138 board, all goes well until I try to stop the debug session and switch back to the editor perspective (by clicking the red square by the pause and run buttons). I have tried disconnecting from the board then switch over. It still crashes. I have also just used the buttons on the upper right to simply step back into the C/C++ perspective, and I can go back and edit, etc. with no problem. When I attempt to debug with the new code, though, ccs tells me that I have to end the previous debug session before I can start a new one. Once I say OK, the system crashes once again. For some reason, every time I attempt to unload the driver for this board, ccsv5 always crashes.

Anybody got any ideas on this one?

  • 0
    TI__Guru**** 171609 points

    Hi,

    First of all, thank you for reporting the compatibility clash between the tools. It is surely useful information for other developers in the future.

    Regarding the problem at hand, the easiest procedure to workaround some of these issues is to follow the procedures at the sections General IDE and Debugger of the CCSv5 Troubleshooting page.

    I regularly use CCSv5.4 with my OMAPL138 experimenter's kit and really couldn't yet find a problem with it. Are you running this on Ubuntu 32/64? What version? Also, what emulator are you using? I wonder if this may be related to the emulator you are using...

    To help you a bit with the development and avoiding terminating the debug session all the time: if you have an active debug session and the project's .out is loaded to the target, anytime a project rebuild is done CCS will ask if you want to reload the code to the target (not terminate, just reload). The same can be achieved if you go to the CCS Debug perspective and go to menu Run --> Load --> Reload Program.

    On the other hand, if you click on the green bug button, the debug session is terminated and restarted. That is the least desirable scenario for you at this moment.

    Hope this helps,

    Rafael

  • 0 in reply to desouza
    Intellectual 955 points

    Yes, I can verify what you are saying. If I have two different programs, a.out and b.out, I can switch back and forth freely by pausing and reloading a different one. The only issue seems to be leaving debug perspective and stopping the debug session with this particular driver.

    BTW, I am using this on Fedora 18 64-bit, ccsv5.4 with all of the latest updates. Thus far this appears to be the sole issue I am experiencing with this setup.

    If you need any further specific information I would, of course, be happy to provide it.

    Thanks,

    Michael

  • 0 in reply to Michael Jordan
    Intellectual 955 points

    I borrowed an SD XDS100v2 and I have tried it on the EXPKITOMAPL138. The results are unchanged. Still crashing.

  • 0 in reply to Michael Jordan
    TI__Guru**** 171609 points

    Michael,

    It may be tied to the Fedora; yesterday I created a VM with this OS and today I am installing CCS (hopefully XDS100 will work on that - not sure given the history).

    I'll let you know if I am successful.

    Cheers,

    Rafael

     

  • 0 in reply to desouza
    Intellectual 955 points

    Thank you for taking the time.

    I just tried running from the command line and I see this when it crashes:

    #
    # An unexpected error has been detected by Java Runtime Environment:
    #
    #  SIGSEGV (0xb) at pc=0x00000000, pid=18227, tid=3000060736
    #
    # Java VM: Java HotSpot(TM) Client VM (11.3-b02 mixed mode linux-x86)
    # Problematic frame:
    # C  0x00000000
    #
    # An error report file with more information is saved as:
    # /home/michael/ti/ccsv5/eclipse/hs_err_pid18227.log
    [thread -143124800 also had an error]
    #
    # If you would like to submit a bug report, please visit:
    #   http://java.sun.com/webapps/bugreport/crash.jsp
    #
    Aborted (core dumped)

    The error log is attached.

    I had reported a similar crash here: http://e2e.ti.com/support/development_tools/code_composer_studio/f/81/p/270567/945354.aspx#945354

    which I did correct by rebuilding my workspace. It seems that this is still happening though only in this one circumstance.

    hs_err_pid18227.log.gz
  • 0 in reply to Michael Jordan
    TI__Guru**** 171609 points

    Michael,

    Unfortunately I could not reproduce this error here in my setup; I am able to terminate a debug session normally from my Fedora 18 VM (same board and emulator).

    One thing that may help you: I installed CCS as root and I am running it as user (I noticed from your log that you probably installed it as user), therefore this may help with this lockup.

    Another detail: my VM is set to a single CPU (yours have two) and my memory is only 1GB (you have 8GB). Of course linking these two parameters to the problem are a long shot, but these were the only differences I could spot.

    Regards,

    Rafael

  • 0 in reply to desouza
    Intellectual 955 points

    Well, thanks for taking the time to check. It's definitely still occurring, but I'll go ahead and close the thread. Maybe next release will change something. It doesn't make it completely unusable, just a bit annoying.

    Thanks,

    Michael