Discrepancy 28335 : Debugger vs Chip

Just sharing the solution with the community:

  //  During initialization:

  EALLOW;

  // GPIO12 after reset is configured as input here

  // Initializing signal qualifications

  GpioCtrlRegs.GPACTRL.bit.QUALPRD1 = 75;   // Time between samples is: 6.66nSec * 2 * 75 = 1 uSec

  GpioCtrlRegs.GPAQSEL1.bit.GPIO12 = 0x2;   // 6 samples in TZ1

  EDIS;

  // DELAY MUST BE ADDED HERE

// Read TZ input while still configured as GPIO input

  if ( GpioDataRegs.GPADAT.bit.GPIO12 == 0) { // check Trip zone at power up

     Report_Error (TripZone);

// Before adding the above delay this was sometimes read as '1' although the hardware signal was '0'

// This had never occur once running from the debugger, just from the chip Flash itself.

// The added delay had solved it

}

  EALLOW;

  GpioCtrlRegs.GPAMUX1.bit.GPIO12 = 0x1;    // Configure GPIO12 as TZ1

  // + configurations of the reset of the Trip Zone registers: TZSEL, TZCTL, ...

  // .....

  // ...

// ...

  EDIS;

 ==================

  It wouldn't be expected that GpioDataRegs.GPADAT.bit.GPIO12 would change its value 

  once a signal qualification is programmed, assuming a stable hardware signal there,

  but the fact is that it need some time to 'relax' after signal qualification programming and before before it is read.