Watchdog for TIVA C

Hello Mohsin

To use the watchdog

SysCtlPeripheralEnable(SYSCTL_PERIPH_WFOG0);

WatchdogReloadSet(WATCHDOG0_BASE,SysCtlClockGet()); // 1 sec watchdog timer

WatchdogIntEnable(WATCHDOG0_BASE);

IntEnable(INT_WATCHDOG);

WatchdogResetEnable(WATCHDOG0_BASE);

WatchdogEnable(WATCHDOG0_BASE);

Please do map the watchdog interrupt handler. On failing to do so 1 sec after interrupt the reset will occur

Regards

Amit

Hello Amit,

Thanks for your reply.

See my code. What I am looking for is that when a certain condition gets true, it should completely reset the processor. Kindly guide me on this.

Regards,

Mohsin

int
main(void)
{

SysCtlClockSet(SYSCTL_SYSDIV_2_5 | SYSCTL_USE_PLL | SYSCTL_OSC_MAIN |
SYSCTL_XTAL_16MHZ);

InitConsole();

UARTprintf("Hello \n");

//

// Enable the peripherals used.

//
SysCtlPeripheralEnable(SYSCTL_PERIPH_WDOG0);

//
// Enable processor interrupts.
//
IntMasterEnable();

//
// Enable the watchdog interrupt.
//
IntEnable(INT_WATCHDOG);

//
// Set the period of the watchdog timer.
//
WatchdogReloadSet(WATCHDOG0_BASE, SysCtlClockGet());

//
// Enable reset generation from the watchdog timer.
//
WatchdogResetEnable(WATCHDOG0_BASE);

//
// Enable the watchdog timer.
//
WatchdogEnable(WATCHDOG0_BASE);
for (;;)
{
// let say

if(condition==true)

{

//reset the processor

}

}

Hello Mohsin,

If the intent is to reset the device under certain condition then I would suggest using SysCtlReset() function instead of using a watchdog.

Regards

Amit

Amit's MCU Reset method surely works - yet another possibility may exist.

if you add an "else" clause to your, "if(condition==true)" and if the watchdog is not reloaded too often, elsewhere w/in your code - then the "else's" containing:

WatchdogReloadSet(WATCHDOG0_BASE, WATCHDOG_RELOAD_VALUE);

will prevent the watchdog from asserting.

Limiting the watchdog's deployment - as I've suggested above - is not, "best practice" - but may be all that's required for simple/limited programs. 

Using "just" the forced MCU SW Reset - as suggested earlier - may leave you vulnerable to other program or system irregularities.   For example - certain program "irregularities" have likely caused your "condition" to be detected - but that same irregularity may be severe enough to "spoil" the condition setting to, "true!"   (note: we've seen this!) 

The watchdog is intended to "escape" your program from such occurrences - it may prove best to combine both the watchdog and a strategic SW Reset - to gain the most robust program operation...