ws2812: add nrf52 support, update timings for m0 and avr

This commit adds support for the nrf52, which is a Cortex-M4 running at
64MHz.
While implementing this, I discovered that the timings for AVR/16MHz and
Cortex-M0 at 16MHz weren't quite right. They do work, but were slightly
out of spec. So I fixed this as well.

ws2812/ws2812_avr_16m.go

@@ -14,6 +14,12 @@ func (d Device) WriteByte(c byte) error {
 	portSet, maskSet := d.Pin.PortMaskSet()
 	portClear, maskClear := d.Pin.PortMaskClear()
 
+	// See:
+	// https://wp.josh.com/2014/05/13/ws2812-neopixels-are-not-so-finicky-once-you-get-to-know-them/
+	// T0H: 4  cycles or 250ns
+	// T0L: 14 cycles or 875ns -> together 18 cycles or 1125ns
+	// T1H: 9  cycles or 562ns
+	// T1L: 8  cycles or 500ns -> together 17 cycles or 1062ns
 	avr.AsmFull(`
 	send_bit:
 		st    {portSet}, {maskSet}     ; [2]   set output high
@@ -21,13 +27,14 @@ func (d Device) WriteByte(c byte) error {
 		brcs  skip_store               ; [1/2] branch if this bit is high (long pulse)
 		st    {portClear}, {maskClear} ; [2]   set output low (short pulse)
 	skip_store:
-		nop                            ; [6]   wait before changing the output again
-		nop
-		nop
+		nop                            ; [4]   wait before changing the output again
 		nop
 		nop
 		nop
 		st    {portClear}, {maskClear} ; [2]   set output low (end of pulse)
+		nop                            ; [3]
+		nop
+		nop
 		subi  {i}, 1                   ; [1]   subtract one (for the loop)
 		brne  send_bit                 ; [1/2] send the next bit, if not at the end of the loop
 	`, map[string]interface{}{