MSP430FR2355: P1IN not capturing???

In what sense does the history "revert" to 0? Are there non-0 entries, but when you breakpoint they've all been cleared? Try setting a breakpoint at the second line of main() just in case there's a surprise reset.

What sort of switches are attached to P1.x? Are they two-position or momentary? The PORT interrupts recognize changes, not levels, in switch state, Also, most physical switches require some form of debouncing.

Hi Bruce...

I will do my best to answer....I have scaled down the project to understand this and attached the complete code (minus function details)..I'd be more than happy to send them as well.

1.  History is zeroed out at the beginning per definition.....In this test code I do a memset after TIMERB0 goes off.  

2.  If I break within TIMER2 at pHistory = history I see P1 changes.  But if I remove this breakpoint and place one within TIMER0 at pHistory = History all changes are gone and I see 0x00 (essentially the pulldowns from IOConfig macro)...If I change to pullups I would get them instead, that is 0xFE (bit 0 is an output).

3. I reverted out of main and into TIMER0 because problem existed in main, so I'm backing up to the source here

4.  Ok I'm using an EVAL board and I have no switches (yes I am aware of debounce).  I am just taking a jumper and tapping it along P1 pins to see them register the change every tenth of a second (ie record history).  I am aware that PORT1 VECTOR is purely Interrupt driven.

5.  Basically what I am trying to do is have a running background (ie TIMER2) constantly surveying pins.  When a P1 IRQ goes off. I want to start the two second timer to validate if any of the pins have been asserted for 2 seconds.  Within main I will look at the history and either pass or fail on any pins that have had a 2 second hold.  I also don't want to lose the other pin status that I've captured that may be under 2 seconds as that has value also....Also once I wake in main and do this task I want to re-enable the switch settings..In this code you do not see me disable the P1 but in TIMER0 most likely I will to freeze the edges captured.

Hope this helps....I'm really struggling with these PxIN....They are not that friendly to use....If I was only concerned with them TIMER2 works fine, but trying to grab edges off of P1 seems to blow everything up

Thanks

Steve

#include <msp430.h>
#include <string.h>
#include <stdint.h>
#include "rc.h"
#include "LPRSradio.h"

volatile unsigned int history[20] = {0}, *pHistory;
volatile char indicator = 0xFE;

int main(void)
{
    WDTCTL = WDTPW | WDTHOLD;   // stop watchdog timer

    pHistory = &history[0];
    IOconfig;
    initClockTo16MHz();
    initTimerB0();
    initTimerB2();
    __bis_SR_register(LPM3_bits + GIE);
}

#pragma vector=TIMER2_B0_VECTOR  //CCIE every tenth of a second
__interrupt void m100SecondTimerB2(void)
//rolling counter...resets after 20 hits or at 2 sec. timer going off
{
    TB2CCR0 += TENTHSEC;
    *pHistory = (unsigned)(P1IN);
    *pHistory &= ~(0x01); //mask unused pins
    pHistory++;
    if (pHistory > history + 19)
        pHistory = history;
}

#pragma vector=TIMER0_B1_VECTOR  //TBIE overflow every 2 seconds
__interrupt void twoSecondTimer(void)
{
    switch(__even_in_range(TB0IV,TB0IV_TBIFG))
    {
      case TB0IV_NONE:   break;               // No interrupt
      case TB0IV_TBCCR1: break;               // CCR1 not used
      case TB0IV_TBCCR2: break;               // CCR2 not used
      case TB0IV_TBIFG:                       // overflow
          pHistory = history;
          memset(history, 0, 20*sizeof(history));
          break;
      default: break;
    }
}

//ISR for on/off buttons
#pragma vector=PORT1_VECTOR
__interrupt void indicatorButtons(void)
{
    TB2CCTL0 |= CCIE;
    launchTimerSeq();
    switch (__even_in_range(P1IV, 16))
    {
        case 4: //P1.1
            indicator ^= SW1;
            P1IFG &= ~BIT1;
            break;
        case 6: //P1.2
            indicator ^= SW2;
            P1IFG &= ~BIT2;
            break;
        case 8: //P1.3
            indicator ^= SW3;
            P1IFG &= ~BIT3;
            break;
        case 10: //P1.4
            indicator ^= SW4;
            P1IFG &= ~BIT4;
            break;
        case 12: //P1.5
            indicator ^= SW5;
            P1IFG &= ~BIT5;
            break;
        case 14: //P1.6
            indicator ^= SW6;
            P1IFG &= ~BIT6;
            break;
        case 16: //P1.7
            indicator ^= ALL;
            P1IFG &= ~BIT7;
            break;
        default:
            break;
    }
}

This looks suspicious. sizeof(history)==(20*2), so this runs far off the end. That may or may not bear on your symptom, but it's a good way to produce random results.

No sir....checked this as I initially had 20 down....turns out memset wants a byte in that field and history is an integer....When I placed a 20 there only half the array gets cleared out.

I recommend:

>         memset(history, 0, sizeof(history));

Fair enough...I will give it a try but it still doesn't solve my problem....That line was put in the code late as an attempt to try things.....Any ideas about the PxIN?

Yes I came upon the fact I was coming into the P1 VECTOR multiple times....which I believe is a big issue .....I've since only allow the timer to launch once ......

thank you for your help.