Hi Experts,
I am trying to connect a third-party JTAG Debugger(Yokogawa Digital Computer adviceLUNA) to Juno-r1 Board.
When using the firmware provided from Recovery-CD, the debugger is connected to Juno.But when using the firmware builded 15.12 LSK with DEBUG options,I can't connect in spite of same procedure and environment.
According to a error message showed by the debugger, the debugger fails to access to the Mem-AP's TAR register.
What should I confirm to connect adviceLUNA to Juno-r1 Board?
Best regards.
Hello,
The fact that it works with the firmware from the recovery CD but not with the latest release suggests that this is caused by CPUIdle, which isn't present in the deliverables provided on the recovery CD.
We have an FAQ regarding this here.
Please can you try running the script provided in that FAQ?
You also mention that you're using Juno R1. Assuming you're using DS-5 to connect to the board, I recommend ensuring that you're using the "Juno ARM Development Platform (r0) | Bare Metal Debug | Debug Cortex-A57/Cortex-A53 big.LITTLE" target. This will allow you to connect to all of the cores (even though it says '(r0)' and you're using R1).
Best regards,
Ash
Thank you for your response.
I tried the method taught by you, but I couldn't connect.
CPUIdle is one of the functions of the Linux Kernel, isn't it?
When using the firmware from the recovery CD, I can connect while UEFI-shell is working before Linux Kernel boot.However, it fails when using the firmware from LSK.
Therefore I'm thinking some factors that disturbs JTAG connect are besides CPUIdle.
What kind of factor is thought about?
Hi,
I'm slightly confused as to exactly when you're unable to connect the debugger.
Please can you clarify what happens when you do each of the following scenarios? Also, please can you erase all flash before trying each scenario. The steps for doing this are as follows:
Cmd> flashFlash> eraseallFlash> quitCmd> usb_on
Cmd> flash
Flash> eraseall
Flash> quit
Cmd> usb_on
Once I understand when it is that you're unable to connect the debugger, there's a better chance that I can help you.
I look forward to hearing from you.
I'm sorry for my poor explanation.
I answer your question.
1.Perform a flash eraseall as outlined above, then flash the recovery CD onto the Juno. Boot to the UEFI shell, then try to connect the debugger. Does this work?
Yes.
2.Perform a flash eraseall as outlined above, then flash the recovery CD onto the Juno. Boot to the Linux shell, then try to connect the debugger. Does this work?
4.Perform a flash eraseall as outlined above, then flash the 15.12 deliverables onto the Juno. Boot to the UEFI shell, then try to connect the debugger. Does this work?
No.
5.Perform a flash eraseall as outlined above, then flash the 15.12 deliverables onto the Juno. Boot to the Linux shell, then try to connect the debugger. Does this work?
6.If the answer to (5) is NO, disable CPUIdle using the script given in my earlier post, then try to connect the debugger. Does this work?
Thanks for clarifying.
It's interesting that you're not even able to connect to the Juno in UEFI when using the latest deliverables. Out of interest, which target are you trying to connect to? My best guess at this point is that you're trying to connect to one of the Cortex-A53 cores, but it's powered down when using the new deliverables as they make one of the Cortex-A57 cores the primary core.
If this is the case, please can you try connecting to:
ARM Development Boards > Juno ARM Development Platform (r1) > Bare Metal Debug > Debug Cortex-A57_0
And see if that works?
I am trying to connect to one of the Cortex-A57 cores.
I confirmed my procedures and configuration files of my debugger again,I wasn't going to connect to the Cortex-A53.
> If this is the case, please can you try connecting to:> ARM Development Boards > Juno ARM Development Platform (r1) > Bare Metal Debug > Debug Cortex-A57_0
Is this the configuration of DS-5?I'm using a third-party debugger, therefore I'm selecting similar thing.
Ah, my bad, I thought you were using a 3rd party JTAG in conjunction with DS-5. I'm afraid I'm not sure what the problem could be then. It's very strange that you're able to connect using the recovery CD, but not even able to connect during UEFI when using the latest deliverables.
Do you have access to a DS-5 installation and a DSTREAM? It would be interesting to see if the same issues are replicated using these tools instead.
Other than that, I really don't know what the issue is. The fact that it works with the recovery CD suggests it's not a hardware problem, but rather a software problem. But, we're not aware of anyone having these issues, at least when using DS-5 with a DSTREAM.
It might be worth contacting the distributor of your debugger (YDC?) to see if they're aware of any similar issues, or to see if they're able to replicate the issue using a Juno with the latest deliverables.
Also, you could try the inverse of what I suggested last time, and try to connect to Core0 of the A53 cluster?
View all questions in Arm Development Platforms forum