LS1012A Custom Board Startup With CodeWarrior TAP

[jclasquin]

Hello, I know similar questions have been posted and I have read them all.  But this seems a little different.

My board is based on LS1012ARDB.  All memory is blank.  When I try to connect the CodeWarrior TAP I see "Failed to write memory at address 0x1ee00e4 on core CortexA53#0.

I have already grounded CLK_OUT to use hardcoded RCW, and I have edited cwflash.py to use safe RCW.  

I have an LS1012ARDB board, so I connected TAP to it and used the flash programmer to erase memory, making it like a new board.  Then I opened a command prompt and followed the instructions to program with aarch64-fsl-gdb.exe.  Everything worked fine.

But my board still gives me the same error about writing 0x1ee00e4.

I turned on logging and compared the messages I get from the reference board and my board.  Attached.

Why do both boards show "LS1043A" in the ccs_config_chain instead of LS1012A?

Why does the reference board show 16 devices in ccs_get_config_chain but my board only shows 3?  Is this why I'm getting the error?

Thank you for any suggestions,

Joel

[jclasquin]

Any suggestions?  Please?

I'd really like to know what the ccs_get_config_chain is telling me.  I assume there is a hardware problem on my board, but I really want to know if the LS1012A is responding (or present) or not.  

Since "LS1043" is shown in the list, is that really the LS1012A?  Since I also see LS1043A on my board, does that mean the chip on my board is at least partially working?

The chain list for the RDB reference board shows LS1043A plus two more "Cortex-A53" entries.  My board does not show any "Cortex-A53" entries.  Is that the problem?  I get an error saying "failed to write... on core CortexA53#0".

I don't think this problem has anything to do with RCW.  I used CodeWarrior's flash programmer to completely erase the QSPI memory on the LS1012ARDB board.  Then I was able to reprogram that board and get it working again because USE_SAFE_RCW is true in my init file.  So I know that works, but not on my board.

Any help would be much appreciated.  Thank you.

[jclasquin]

I should have mentioned that when I was using the command prompt I was following Yiping's directions here...

https://community.nxp.com/t5/CodeWarrior-for-QorIQ/LS1012-custom-board-JTAG-speed-too-low/m-p/600758...

That didn't work, so I tried to connect in the CodeWarrior IDE.  Same error.  Then I tried "Diagnose Connection" and turned on logging.  That's when I noticed the differences in the log files.

Again, thank you.

Joel

[jclasquin · Accepted solution]

I found the problem.  The LS1012A was being held in reset because of SDA_RST_TGTMCU.  That signal is pulled up to the 3.3V supply derived from USB Vbus.  It was floating without the USB cable plugged in.  Now the LS1012A's reset line is high, it shows "Cortex-A53" devices in the list and I'm able to connect the TAP and program the part.

Joel