Stack errors with TI-RTOS and Tiva C launchpad

Dale,

The first thing that comes to mind is that you might be overflowing your tasks' stack. When the program fails can you open ROV and send me a snapshot of the Detailed view of the Task module? Also are you using the TI-RTOS I2C drivers? If you are, what version of TI-RTOS are you using?

Thanks,

Moses

Hi Moses

Here is the screen shot of the ROV Detailed Task module

I am using TI-RTOS for Tiva C version 2.1.0.03. I am using the TI-RTOS I2C drivers.

4035.Task Module.docx

Thank you

Dale

Dale,

 From ROV your  task stack looks healthy. I still don't have enough information to narrow this down. I see from ROV that the DataGather task was running before the abort. What are you doing in that task? Where in that task does it fail? Can I see how you setup the I2C driver? Your cfg file will also be useful to help me narrow the problem down.

There's also a chance that you might be corrupting the task stack pointer. Are using any heap?

Thanks,
Moses

Hi Moses:

The purpose of DataGather() is to read data from the MEMS accelerometer and ADCs when data is available. 

I have included main.c and app.cfg, the main.c contains the I2C driver configuration as well as DataGather().

Examining the list of functions that led to the abort, DataGather is not on the list, but MEMSIntHandler is where it seems to interrupt. This function is also in main.c. The purpose of this interrupt handler is to post a semaphore (MEMSReady) when the data from the MEMS sensor is ready. The MEMS sensor is configured to send an interrupt back to the MCU when data is ready. This interrupt appears at the MCU at GPIO Port A Pin 6. The steps ideally I would eventually like to get into the DataGather() routine are as follows:

1. A request is made to read 6 bytes of data from the MEMS. This 6 bytes will be X, Y and Z axis acceleration values.
2. The request will pend on the 'MEMSReady' semaphore.
3. When the data is read, DataGather will then pend on a semaphore indicating that the ADC data is available.
4. DataGather will then read the ADC data.
5. All these data points will be placed in an array for use by the control routine.

You will see in my file I commented out the ADC statements, to see if that could be the cause of the error.

I have not configured any heap, unless it has been configured by RTOS.

Let me know what other information I can provide.

If necessary, I can export the project and send the archive file.

Thank you

Dale

3404.main.c

8715.app.cfg

Dale,

First thing I see is that you're creating a Hwi for the GPIO interrupt in your cfg file. There's no need for that as the GPIO_setupCallbacks internally creates a Hwi for the GPIO. I don't see the code where you set up the I2C and GPIO pins, I can assume it's alright since at least you're able to get interrupts. Fix this and test it.

Regarding the heap, in your *cfg file I see BIOS.heapSize is set to 0. So if you're doing any dynamic module creates or any memory allocation, then that will be an issue.

I'll need more information. You said the program runs and aborts in the MEMSIntHandler. Is it the first time it runs that ISR that it aborts? Can I see what the console window displays when its aborts (abort message)? When it aborts can you open ROV, go to the BIOS module and go to the Scan for errors tab. I'd like to see what error is displayed there.

Thanks,

Moses