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.

CCS support for OMAP2430 on Windows 8 ???



I have a customer with the following question about an older OMAP processor and how they can get a version of CCS to run on Windows 8, can someone help?  Thank you!

We have recently upgraded our computers and have run into an unexpected issue. It seems that the OMAP we are using (a variant of the 2430) is NOT supported by any version of Code Composer newer than 3.3. But CCS3.3 is very difficult, if not impossible, to get running on Windows8, which is the only OS now available on modern computers. So this leaves us in a rather difficult and precarious situation, being unable to develop software for our hardware platform. What do you suggest? Thanks!

  • Hi Kevin,

    You have two options:

    1) use CCSv3.3 on an XP virtual machine

    2) manually create a target configuration for OMAP2430 in CCSv5. I looked at the layout in CCSv3.3 and you should be able to create one for CCSv5 manually. I can create it for you if you tell me which emulator you will use. But because that device is not officially supported by v5, there is no guarantee it will work

    Thanks

    ki

  • Thanks for the quick reply Ki.  Would you mind helping my customer create a custom target?  Here was their response to our thread:

    "Thanks for the info. I tried to create a target config manually, but could not get it to work. Maybe I did something wrong. Would be very happy to try one created by someone who knows what they are doing. The emulator is a Blackhawk USB560M (BH-USB-560M) and the device is the OMAP2531. I created a profile with the IcePickB router and 3 devices: ARM11, bypass, C64XP. Could run the JTAG test successfully, but could not connect to the DSP. Don’t really care about the ARM or need to connect to it, but do need to connect to DSP and be able to load code, set breakpoints, view memory, etc.

    I do recall a long time ago tinkering with CCS3.3 on an XP VM, but that was really tricky. We prefer to use Virtual Box because it is free and works well for our linux VMs, but I could never get CCS to work under Virtual Box. I was eventually able to get it work in VMWare, but it was a real chore as I recall. Don’t really want to go down that road, and unfortunately, have already purchased CCS 5.4."

  • Actually, sounds like they created a correct file themselves, especially if the Test Connection utility passed all tests. Are they using the same gel files as was in CCSv3.3? Usually for OMAP devices you need to pull the DSP put of reset and it is some ARM gel file that has the routine to do this.

  • I am the customer Kevin was referring to. I created the file, and CCS did let me into the debug mode without error, but when I right-click the DSP and try to connect, it failed and gave me an error code that said the device ID was not recognized. If I ran the "test connection" function, all tests passed. So something is not right and I really don't know what.

    In looking at my old CCS3.3 setup on a WinXP laptop, I do note that there are lots of OMAP2430-specific files, drivers, gel files, etc. But I don't know which ones to grab, nor do I know what to do with them on the CCS5.4 side.

  • Can you connect to the ARM? The "device ID no recognized" error concerns me. It doesn't sound like an initialization error.

    James Gilley said:
    But I don't know which ones to grab, nor do I know what to do with them on the CCS5.4 side.

    CCSv5 ships with some, but not all, of the OMAP2430 gel files. It has the main one for the ARM however (.\ccsv5\ccs_base\emulation\gel\omap2430_arm11.gel). You can use this as the initialization script for the ARM in your target configuration file. If you open that file in the text editor, you can see that it loads a bunch other GEL files. You need to make sure all those files exist in the same location as the main GEL file. Copy over the gel files that are in v3 but missing in the v5 directory.

    Thanks

    ki

  • Now I cannot connect to the ARM, or even launch my configuration, which is odd because I could do it before when I got the device ID failure. I moved the gel files as you instructed and they are now all there. I also went into my custom configuration and compared it to the one on CCS3.3. They are as close to identical as I can make them.

    What happens now when I try to launch my configuration is that I get an error saying ARM11_0: TMS470RXX_regids.xml does not map the register REG_CP15_REG0_TCMS_ID defined in C:\ti\ccsv5\ccs_base\common\targetdb\Modules\ARM_R2X_NotVisible.xml

    So I have no idea what that is all about. But here is what is odd. If I bypass the ARM_11 and then launch, it goes into debug mode without error and shows me the C64XP. But when I right-click it and select connect target, I get this error: IcePick_B_0: Error connecting to the target: (Error -233 @ 0x0) The JTAG IR and DR scan-paths cannot circulate bits, they may be broken. An attempt to scan the JTAG scan-path has failed. The target's JTAG scan-path appears to be broken with a stuck-at-ones or stuck-at-zero fault. (Emulation package 5.1.73.0) , which is different from the error I got a couple of days ago. I also get this JTAG error when running the test connection, which I did not get the last time (which is when I got the device ID error).

    So I'm not sure what to do next. Please advise. Thanks!

  • James Gilley said:
    Error connecting to the target: (Error -233 @ 0x0) The JTAG IR and DR scan-paths cannot circulate bits, they may be broken. An attempt to scan the JTAG scan-path has failed. The target's JTAG scan-path appears to be broken with a stuck-at-ones or stuck-at-zero fault. (Emulation package 5.1.73.0)

    This is a serious error. Usually a HW issue - the JTAG scan chain is bad. It can't successfully circulate bits. If you take the ARM out of bypass, do you still get the error?

    I'm afraid that I don't have much more advice to give. I don't have your target available to me to play with and CCSv5 doesn't officially support OMAP24xx (ex. never tested with it or have device files for it). If you can still get your board to work on 3.3, a Virtual Machine may be your best bet.

  • If I recall correctly, and the OMAP253x documentation should substantiate this claim, the DSP is held in reset by default.  I suspect the series of GELs files that were referenced before actually performed some operations to take the DSP out of reset and hold it in an idle loop until the user then instructs CCS to connect to the DSP.

    The first thing to get established is the connection to the ARM11.  Once we establish that, we then know the JTAG scan path is correct.  At that point we start working towards getting the DSP connected.

  • Okay, so I can now connect once again (I did have a hardware problem with the emulator, which is why it took me so long to post this reply). However, now when I connect, I still get this error, which means nothing to me:

    ARM11_0: TMS470RXX_regids.xml does not map the register REG_CP15_REG0_TCMS_ID defined in C:\ti\ccsv5\ccs_base\common\targetdb\Modules\ARM_R2X_NotVisible.xml

    In CCS3.3, I can connect and debug successfully, but only on the DSP side. I was never able to do anything with the ARM. CCS3.3 gets stuck in an infinite loop if you try to connect to the ARM, but connects to the DSP just fine.

    If I do a 'test connection' from my target configuration view, everything passes with no errors, so that is good. But I'm still unable to make CCS5.4 do anything useful.

  • This thread was taken off-line and addressed.

    The issue is now closed.

    For those following the thread, there were two bugs, the first was an error with one of the cpu xml files (causing the first issue). The second was an issue with the driver for the DSP, a conflict with the outdated ICEPICK B router that is on the OMAP24xx. Custom fixes for both addressed the issue.

    Thanks

    ki