TM4C1294NCPDT: Debugging hard faults with CCS

Sasha Gold

Part Number: TM4C1294NCPDT
Other Parts Discussed in Thread: EK-TM4C1294XL

Hello,

I'm encountering a number of varying hard faults in my code, and am looking for advice on how to use Code Composer Studio to its fullest to help troubleshoot these issues.

I'm using:

CCS v11.2.0

XDS100v3

I also have available a Tiva C Series Launchpad EK-TM4C1294XL Rev D

I have read this: https://www.ti.com/lit/an/spma043/spma043.pdf

But unfortunately I keep getting stuck because the Program Counter (PC) Address in Disassembly just points to the FaultISR(). Also the NVIC Fault Status register shows it is a precise data bus error and the Bus Fault Address is valid, however the address points to 0x00000000, which as indicated by the Memory Map in the datasheet is in On-Chip Flash.

As I understand it, if the Fault ISR is more than a simple while loop, it may require an offset to the PC to find the offending instruction, but my code is just a simple while loop:

static void
FaultISR(void)
{
    //
    // Enter an infinite loop.
    //
    while(1)
    {
    }
}

Adding some of my debugging screenshots in case it helps:

Any help as to how to actually find the offending instruction is greatly appreciated.

over 2 years ago

0 Charles Tsai over 2 years ago

TI__Guru**** 191486 points

Hi,

A lot of time the fault is due to two main reasons. First, you try to access a peripheral that is not yet enabled. For example, see the below code. Here I try to access TIMER0 registers without first enable TIMER0 because I comment out SysCtlPeripheralEnable(SYSCTL_PERIPH_TIMER0). This is often the first thing to check. The system will trip a bus fault when accessing a peripheral without first enabling it.

//
// Enable the peripherals used by this example.
//
//SysCtlPeripheralEnable(SYSCTL_PERIPH_TIMER0);

//
// Configure the two 32-bit periodic timers.
//
TimerConfigure(TIMER0_BASE, TIMER_CFG_PERIODIC);

The next type of reason is that you don't have enough stack or heap (if you use heap for dynamic memory allocation). Most often, it is due to not enough stack. So increase the stack to see if the problem goes away. If you have a recursive calls in a function then it can easily go beyond stack quickly.

0 Sasha Gold over 2 years ago in reply to Charles Tsai

Prodigy 143 points

Thanks for your reply Charles.

I have checked for stack and heap overflow, I seem to have plenty to spare. I have also tried increasing the stack size just in case.

As far as I can tell I am not using any peripheral before enabling it.

I'm thinking the cause(s) is less obvious, since the hard fault seems to occur randomly, sometimes before and sometimes after peripheral initialization.

Is there anything I can do to make the stack pointer point me in the right direction?

0 Sasha Gold over 2 years ago in reply to Sasha Gold

Prodigy 143 points

For more background:

I used the method here to check for stack overflow: https://e2e.ti.com/support/microcontrollers/arm-based-microcontrollers-group/arm-based-microcontrollers/f/arm-based-microcontrollers-forum/1164811/tm4c1230e6pm-reset-of-micro/4386491#4386491

And used the vPortGetHeapStats() provided by FreeRTOS to check heap.

0 Charles Tsai over 2 years ago in reply to Sasha Gold

TI__Guru**** 191486 points

Hi,

I'm not sure what is the cause of your problem?

- Is your application based on FreeRTOS?

- Can you repeat the same fault in another board you have?

- I don't know how complicated is your application. Can you somehow simplify your code by removing some portions and gradually add back the removed portions. This way it is easier to debug. Your log seems to indicate it is a bus fault with precise address but the address is 0 which I have not seen before.

- Perhaps you want to insert some log_info at different places of code to have a better assessment on where and when the fault is coming

0 Sasha Gold over 2 years ago in reply to Charles Tsai

Prodigy 143 points

Hi Charles,

I am not expecting anyone to find the source of my problem, I was just hoping to find more information on how to use the Program Counter and NVIC registers for help with debugging.

I am currently attempting the rudimentary troubleshooting methods you mentioned, but I was hoping CCS might have some tricks up its sleeve to assist in this.

Can you tell me, if the PC points outside of the stack, is it essentially useless?

Do you know if there's any way to get more hardware tracing with the TM4C and CCS? Would a different debugger tool than the XDS100v2 help?

This post mentions something about gel scripts, is there more information about that somewhere? Could I use this method in my case?

https://e2e.ti.com/support/tools/code-composer-studio-group/ccs/f/code-composer-studio-forum/810598/ccs-tm4c1294kcpdt-how-do-i-get-the-stack-unwound-in-exception-handlers/3026599#3026599

Thanks again,

0 Sasha Gold over 2 years ago in reply to Sasha Gold

Prodigy 143 points

I read here that 7 years ago TI was working on providing an app note for debug and trace, do you know if that app note exists now and where I can find it?

e2e.ti.com/.../hardware-trace-on-tiva-series-mcu-s

0 Charles Tsai over 2 years ago in reply to Sasha Gold

TI__Guru**** 191486 points

Sasha Gold said:
This post mentions something about gel scripts, is there more information about that somewhere? Could I use this method in my case?

https://e2e.ti.com/support/tools/code-composer-studio-group/ccs/f/code-composer-studio-forum/810598/ccs-tm4c1294kcpdt-how-do-i-get-the-stack-unwound-in-exception-handlers/3026599#3026599

HI,

The post you refer to authored by Chester is very useful. I actually try it myself for debugging faults. Follow his instructions and see if you can have a better idea on the offending instructions or functions by looking the callstack.

0 Charles Tsai over 2 years ago in reply to Charles Tsai

TI__Guru**** 191486 points

Hi Sasha,

If your fault problem resolved? I haven't heard back from you. I will close the thread for now. If you have some updates, you can write back to this thread and it will automatically reopen.

Arm-based microcontrollers

Arm-based microcontrollers forum

TM4C1294NCPDT: Debugging hard faults with CCS