Not-so-Tiny-Dimmer

/*
 * Core driver routines for WS2812 LED strips 
 * This is a stripped down and modified version of cpldcpu's light_ws2812 library.
 * 
 * Author: Tim (cpldcpu@gmail.com), Patrick Pedersen (ctx.xda@gmail.com)
 *
 * License: GNU GPL v2+ (see License.txt)
 */

#include <avr/interrupt.h>
#include <avr/io.h>
#include <util/delay.h>

#include "config.h"
#include "ws2812.h"
 
#if STRIP_TYPE == WS2812

/*
  This routine writes an array of bytes with RGB values to the Dataout pin
  using the fast 800kHz clockless WS2811/2812 protocol.
*/

// Timing in ns
#define w_zeropulse   350
#define w_onepulse    900
#define w_totalperiod 1250

// Fixed cycles used by the inner loop
#define w_fixedlow    3
#define w_fixedhigh   6
#define w_fixedtotal  10   

// Insert NOPs to match the timing, if possible
#define w_zerocycles    (((F_CPU/1000)*w_zeropulse          )/1000000)
#define w_onecycles     (((F_CPU/1000)*w_onepulse    +500000)/1000000)
#define w_totalcycles   (((F_CPU/1000)*w_totalperiod +500000)/1000000)

// w1 - nops between rising edge and falling edge - low
#define w1 (w_zerocycles-w_fixedlow)
// w2   nops between fe low and fe high
#define w2 (w_onecycles-w_fixedhigh-w1)
// w3   nops to complete loop
#define w3 (w_totalcycles-w_fixedtotal-w1-w2)

#if w1>0
  #define w1_nops w1
#else
  #define w1_nops  0
#endif

// The only critical timing parameter is the minimum pulse length of the "0"
// Warn or throw error if this timing can not be met with current F_CPU settings.
#define w_lowtime ((w1_nops+w_fixedlow)*1000000)/(F_CPU/1000)
#if w_lowtime>550
   #error "Light_ws2812: Sorry, the clock speed is too low. Did you set F_CPU correctly?"
#elif w_lowtime>450
   #warning "Light_ws2812: The timing is critical and may only work on WS2812B, not on WS2812(S)."
   #warning "Please consider a higher clockspeed, if possible"
#endif   

#if w2>0
#define w2_nops w2
#else
#define w2_nops  0
#endif

#if w3>0
#define w3_nops w3
#else
#define w3_nops  0
#endif

#define w_nop1  "nop      \n\t"
#define w_nop2  "rjmp .+0 \n\t"
#define w_nop4  w_nop2 w_nop2
#define w_nop8  w_nop4 w_nop4
#define w_nop16 w_nop8 w_nop8

static uint8_t _sreg_prev, _maskhi, _masklo;

/* ws2812_prep_tx
 * --------------
 * Description:
 *      Prepares for a data transmission to the WS2812 strip.
 *      Always call this function before calling ws2812_tx_byte()!
 */
void ws2812_prep_tx()
{
        _masklo = ~WS2812_DIN_MSK & WS2812_DIN_PORT;
        _maskhi = WS2812_DIN_MSK | WS2812_DIN_PORT;

        _sreg_prev=SREG;
        cli();  
}

/* ws2812_wait_rst
 * ---------------
 * Description:
 *      Waits for the WS2812 to reset.
 */
void ws2812_wait_rst()
{
#if defined(WS2812_RESET_TIME) && WS2812_RESET_TIME > 0
        _delay_us(WS2812_RESET_TIME);
#endif
}

/* ws2812_end_tx
 * -------------
 * Description:
 *      Ends data transmission with the WS2812 by 
 *      restoring the status register to its previous state
 *      and re-enabling interrupts. Always call this function
 *      after data transmission is complete!
 */
void ws2812_end_tx()
{
        SREG=_sreg_prev;
        ws2812_wait_rst();
        sei();
}

/* ws2812_tx_byte
 * --------------
 * Description:
 *      Transmitts a byte of data to the WS2812.
 *      Toggling this function three times consecutively, sets a pixel.
 *      After every third consecutive transmission, the next pixel is
 *      written to. Transmissions MUST take place consecutively, else
 *      there's a risk that the timing fails and the strip will either
 *      be written to incorrectly, or not be written to at all.
 */

#pragma GCC push_options
#pragma GCC optimize ("O0")

void ws2812_tx_byte(uint8_t data)
{
        uint8_t ctr;

        asm volatile(
                "       ldi   %0,8  \n\t"
                "loop%=:            \n\t"
                "       st    X,%3 \n\t"    //  '1' [02] '0' [02] - re
                #if (w1_nops&1)
                w_nop1
                #endif
                #if (w1_nops&2)
                w_nop2
                #endif
                #if (w1_nops&4)
                w_nop4
                #endif
                #if (w1_nops&8)
                w_nop8
                #endif
                #if (w1_nops&16)
                w_nop16
                #endif
                "       sbrs  %1,7  \n\t"    //  '1' [04] '0' [03]
                "       st    X,%4 \n\t"     //  '1' [--] '0' [05] - fe-low
                "       lsl   %1    \n\t"    //  '1' [05] '0' [06]
                #if (w2_nops&1)
                w_nop1
                #endif
                #if (w2_nops&2)
                w_nop2
                #endif
                #if (w2_nops&4)
                w_nop4
                #endif
                #if (w2_nops&8)
                w_nop8
                #endif
                #if (w2_nops&16)
                w_nop16 
                #endif
                "       brcc skipone%= \n\t"    //  '1' [+1] '0' [+2] - 
                "       st   X,%4      \n\t"    //  '1' [+3] '0' [--] - fe-high
                "skipone%=:               "     //  '1' [+3] '0' [+2] - 

                #if (w3_nops&1)
                w_nop1
                #endif
                #if (w3_nops&2)
                w_nop2
                #endif
                #if (w3_nops&4)
                w_nop4
                #endif
                #if (w3_nops&8)
                w_nop8
                #endif
                #if (w3_nops&16)
                w_nop16
                #endif
                "       dec   %0    \n\t"    //  '1' [+4] '0' [+3]
                "       brne  loop%=\n\t"    //  '1' [+5] '0' [+4]
                :	"=&d" (ctr)
                :	"r" (data), "x" ((uint8_t *) &WS2812_DIN_PORT), "r" (_maskhi), "r" (_masklo)
        );
}

#pragma GCC pop_options

#endif