clock16.h

// -*- mode: c++; indent-tabs-mode: nil; -*-

#ifndef ARDUINO_MINUS_MINUS_CLOCK16_H
#define ARDUINO_MINUS_MINUS_CLOCK16_H

#ifdef ARDUINO_MINUS_MINUS_CLOCK32_H
#error "Only one clock resolution can be used"
#endif

/* This file should be included - once - in the main C++ file of the 
   application.

   Other headers should - if at all possible - not include this file and
   use a template instead.
 */

/** This is a Clock with 16bit clock resolution.
    
    The value from Clock16::millis() will wrap around after about 65 seconds.

    The recommended way to use a clock value in user code is:

    typename Clock16::time_res_t now = Arduino16::millis();

    Motivation: With Clock16 instead of Clock32, Lars has seen a code size 
    reduction of 238 bytes with avr-gcc 4.6.1.
 */

// Define this for a slower clock (for low power modes). Note that you should
// also set TIMER0_MICRO_SCALE if you don't use the default here
// See the Makefile and timerscale.py for details
#ifndef TIMER0_PRESCALE
# define TIMER0_PRESCALE 64
#endif

typedef _Clock<uint16_t,Timer0> Clock16;

Clock16 clock;

static void clock16_isr() 
    {
    // copy these to local variables so they can be stored in registers
    // (volatile variables must be read from memory on every access)
    typename Clock16::time_res_t m = Clock16::timer_millis;
    uint16_t f = Clock16::timer_fract;

    m += (TIMER0_PRESCALE * (256 / (F_CPU / 1000000))) / 1000;
    f += (TIMER0_PRESCALE * (256 / (F_CPU / 1000000))) % 1000;
    if (f >= 1000)
        {
        f -= 1000;
        m += 1;
        }

    Clock16::timer_fract = f;
    Clock16::timer_millis = m;
    Clock16::timer_overflow_count++;
    }

/* 
 * Implementation of the clock ISR for 16 bits resolution 
 */
#ifndef CLOCK_NO_ISR
ISR(TIMER0_OVF_vect)
    {
    clock16_isr();
    }
#endif

#endif