Forum Discussion

Altera_Forum's avatar
Altera_Forum
Icon for Honored Contributor rankHonored Contributor
11 years ago

Arm DS5 debugger Issue

I am trying to load a hello world program directly into the on chip RAM of the altera Cyclone 5 sockit. The first time I did this some time ago, I did not seem to have an issue. However, since then every time i try to debug one of the cores, eclipse tells me : Cannot connect to TCP:localhost. I am wondering if I am just doing something stupid or what. I have tried a debugger script which does nothing. It seems like eclipse cannot connect to localhost. I have no firewalls.

Any help is appreciated.

11 Replies

  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    11 years ago

    Go under run-Debug configurations and make sure you have the right jtag interface selected and that you have drivers for it

  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    11 years ago

    Thanks for the response.

    I am fairly confident I have things setup correctly. In device manager, under jtag cables I can see CV sockit jtag and serial console. The device installed the drivers correctly when I powered the sockit on. Additionally, under target connection, I have USB blaster selected and for bare metal debug: CV sockit on localhost USB1. I think that if things were mis configured in this area, I would not get the bare metal debug option that pops up with the CV sockit connection.
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    11 years ago

    Some more information after digging around for awhile. If I click on one of the errors it says "debug_server.py failed during execution. Port 3010 already in use. This would indicate that gdb is trying to use port 3010 and is blocked. However, it's not blocked to my knowledge! If I run netstat, I can see that port 3010 does not appear to be in use.

  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    11 years ago

    I had a similar problem when my DS-5 instance was crashing frequently due to an Eclipse bug. When I restarted DS-5 I would receive the error you posted. A Google search for that error came up with this solution:

    http://www.altera.com/support/kdb/solutions/fb145963.html

    Which was not helpful because I knew I didn't have an instance running. After some experimentation with "ps -ef" I discovered that my problem was the "vstrm_serverd_rddi" process still running in the background. I manually killed the process and I was immediately able to reconnect to my debug target. Try that and see if it helps.
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    11 years ago

    mborden: Thanks for the reply

    I hate to say it, but since I was not able to fix the bug at the time, I have since moved to xilinx Zynq :P. It is a shame, because I really liked the Cyclone SOC. The xilinx tools are completely free for the Zynq processor, and the SOCs have similar functionalities... so rather than bang my head against the wall...

    The fix you suggested could work; however, I have to say, that I won't be trying it at the moment. I have since then uninstalled all Altera tools due to hard drive space. If I end up coming back to the Cyclone SOC, I will give your suggestion a shot.

    Thanks
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    8 years ago

    I am facing the same problem. "Cannot connect to TCP:localhost"

    I have tried to resolve it by minimizing the jtag frequency but still it's having the same problem.

    Can anyone have a solution for it??
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    8 years ago

    I get this message sometimes when the CPU crashes. It can mean that the debugger can't contact the CPU. Did you try to reset and/or to power cycle your board? Are you sure its working correctly and that the HPS gets a valid clock input? Did you install a preloader and does it run? Did you check the boot_sel, clk_sel and HPS reset pins?

  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    8 years ago

    Thanks for your valuable reply Daixiwen.

    Yes, I have tried by resetting it and making it on and off many a times.

    And I have verified the HPS by checking the clocks at SDCARD on power on means HPS clocks out to the SDCARD to boot.

    Another thing is that I haven't get any messages or data on console UART on power on HPS.

    Thanks and regards