Santhi,

In fact, this might not prove that there is no software fault, it just might prove that the software reset is flawed and does not properly reset the internal flags.

Please do the following experiment:

• set up the extint1 interrupt so that upon a given stimulus - say, grounding a specific input - it performs ret rather than reti
  
• check, that after powerup extint1 performs as expected whereas after the grounding of the said input it does not anymore
  
• perform the software reset
  
• check whether extint1 execution was resumed or not, and tell us the result.
  

Don't get me wrong: I don't say the software IS flawed, I just say that this might not rule it out.

JW