Problem after update to ccs 5.2 with debugging am335x

Hi Michael,

a couple things

- how and when did you update to 5.2?  From which version?  5.1.1?  5.1.0?  There have been some know issues around updating.

Are you using the same workspaces/projects or importing?

Best Regards,

Lisa

Michael,

There are some options that control what happens after a program load. By default, CCS will set the PC to the program entry point, set a breakpoint at main(), and then attempt to run to main(). It sounds like you have the run to main option set, but your program is not making it to main(). You could try turning the run-to-main option off, and having a look at why the program doesn't make it there by stepping through the initialization code.

A couple of other thoughts: to ensure your program load loaded properly, you can also turn on "full verification" in the debugger options. It might also be worthwhile to check whether the 5.2 update changed the GEL file. Perhaps your program was relying on something that is configured through the GEL initialization script.

Hey,

I have updated from CCS5.1.1 and working with the same workspaces, I had not to import the projects.

Before the update the debugger was at the beginning of the main, and waited for "run". I will search and try this with the run to main option.

The gel file is the same, this works as before.

I have disabled

"run to symbol: main."

But the same happens, after loading(which is unbelievable fast) the core is running. If try to stop this error message is in the console

CortxA8: Trouble Halting Target CPU: (Error -2062 @ 0x1FE) Unable to halt device. Reset the device, and retry the operation. If error persists, confirm configuration, power-cycle the board, and/or try more reliable JTAG settings (e.g. lower TCLK). (Emulation package 5.0.710.0)

I am uncertain how the cortex gets into the running state if the run-to-main option is off.  If you can generate a Debug Server log and post it here, it might help (Go to help-> CCS Support, enable the debug server log, perform the load, zip the log and post it here).

I have enabeld the run to main before I start this session. 3683.debug.zip

Thanks

Thank you. As you were saying the cortex is already running at the end of the program load even with the run-to-main option off, it would be useful to look at the log for this case too. The log with the option on looks normal at least from the perspective that all the expected operations are there.

Is there a fast solution or should I uninstall ccs5.2 and install 5.1.1?

run to symbol is disabled in this log.

5482.debug2.zip

It is only sometimes debugable. This error message occurs after connecting and disconnecting the board from power supply.

IcePick_D_0: Error: (Error -150 @ 0x0) One of the FTDI driver functions used during configuration returned a invalid status or an error. (Emulation package 5.0.681.0) CS_DAP_DebugSS: Error: (Error -154 @ 0xFFFFFF66) One of the FTDI driver functions used to write data returned bad status or an error. (Emulation package 0.0.0.0) CortxA8: JTAG Communication Error CortxA8: Failed to remove the debug state from the target before disconnecting. There may still be breakpoint op-codes embedded in program memory. It is recommended that you reset the emulator before you connect and reload your program before you continue debugging CS_DAP_DebugSS: Failed to remove the debug state from the target before disconnecting. There may still be breakpoint op-codes embedded in program memory. It is recommended that you reset the emulator before you connect and reload your program before you continue debugging IcePick_D_0: Failed to remove the debug state from the target before disconnecting. There may still be breakpoint op-codes embedded in program memory. It is recommended that you reset the emulator before you connect and reload your program before you continue debugging

I will look through the log.

If 5.1.1 is working for you and you want to use it to get started, there is no problem installing v5.1.1 into a different folder and leaving v5.2 alone (assuming yo have enough disk space).  For example, if v5.2 is installed into c:\ti, you can install v5.1.1 into c:\CCS_511.

The log file is empty. Could you repost?

Ok, thanks.

oh, sorry.

0218.debug2.zip

The thing we notice from the log is that it looks like the "Run to symbol" edit box has been cleared, but the two check boxes below are still set ("On a program load or restart", and "On a reset"). The two check boxes need to be unchecked, or CCS will still do a run. Sorry, I should have been more explicit about how to turn those options off.

Since the program seems to load OK, my working theory is there may be something going on in the program that causes the target hardware to get into a bad state. If this is the case, the same problem would have existed in v5.1.1 though. I'm not sure how to explain why it would have been working in v5.1.1.  I was trying to take the approach of letting the program load and then stepping through the initialization code until it was obvious where the the target wasn't behaving as expected. Then, using that clue we could determine the next steps.  So, with the above options turned off, could you give stepping through the initialization code a try?

I have tried this and with the defaults. I`m using two sample applications, so the project properties should be ok. I`m not able to reach a breakpoint. The only thing that is exactly the same by both projects is the target configuration, but I can`t find a wrong configuration.

I`m trying to install ccs5.1.1 in a different location, but the first attempt fails

Most time the session fails, but I have loaded and executed the gel file by me.maybee you can find a difference in the log file from this session, that was succesful3108.debug3.zip.

Ok. It seems not to be a problem of ccs5.2 I have tried to flash a new release on the controller, which fails, but now no application is executed and it is possible to start and execute a debug session. But I dont know why there is no error message an the load of the programm is successful.