From 00a28cd5029420d363dfa55c505caaf11568d81e Mon Sep 17 00:00:00 2001
From: Phil Hord <phord@purestorage.com>
Date: Mon, 21 Nov 2016 17:34:52 -0800
Subject: [PATCH] DDA: use bitmask to track active axes for faster dda_step().

In dda_step instead of checking our 32-bit-wide delta[n] value,
just check a single bit in an 8-bit field.  Should be a tad faster.
It does make the code larger, but also about 10% faster, I think.

Performance:

  ATmega sizes               '168   '328(P)   '644(P)     '1280
  Program:  19696 bytes      138%       65%       32%       16%
     Data:   2191 bytes      214%      107%       54%       27%
   EEPROM:     32 bytes        4%        2%        2%        1%

  short-moves.gcode statistics:
  LED on occurences: 888.
  LED on time minimum: 263 clock cycles.
  LED on time maximum: 532 clock cycles.
  LED on time average: 269.273 clock cycles.

  smooth-curves.gcode statistics:
  LED on occurences: 23648.
  LED on time minimum: 255 clock cycles.
  LED on time maximum: 571 clock cycles.
  LED on time average: 297.792 clock cycles.

  triangle-odd.gcode statistics:
  LED on occurences: 1636.
  LED on time minimum: 255 clock cycles.
  LED on time maximum: 522 clock cycles.
  LED on time average: 283.861 clock cycles.
---
 dda.c | 14 ++++++++++----
 dda.h |  1 +
 2 files changed, 11 insertions(+), 4 deletions(-)

diff --git a/dda.c b/dda.c
index be43087..65dae85 100644
--- a/dda.c
+++ b/dda.c
@@ -187,6 +187,7 @@ void dda_create(DDA *dda, const TARGET *target) {
     dda->id = idcnt++;
   #endif
 
+  dda->active_axes = 0;
   // Handle bot axes. They're subject to kinematics considerations.
   code_axes_to_stepper_axes(&startpoint, target, delta_um, steps);
   for (i = X; i < E; i++) {
@@ -194,6 +195,8 @@ void dda_create(DDA *dda, const TARGET *target) {
 
     delta_steps = steps[i] - startpoint_steps.axis[i];
     dda->delta[i] = (uint32_t)labs(delta_steps);
+    if (delta_steps)
+      dda->active_axes |= 1 << i;
     startpoint_steps.axis[i] = steps[i];
 
     set_direction(dda, i, delta_steps);
@@ -255,6 +258,9 @@ void dda_create(DDA *dda, const TARGET *target) {
     dda->e_direction = (target->axis[E] >= 0)?1:0;
 	}
 
+  if (dda->delta[E])
+    dda->active_axes |= 1 << E;
+
 	if (DEBUG_DDA && (debug_flags & DEBUG_DDA))
     sersendf_P(PSTR("[%ld,%ld,%ld,%ld]"),
                target->axis[X] - startpoint.axis[X], target->axis[Y] - startpoint.axis[Y],
@@ -566,28 +572,28 @@ void dda_start(DDA *dda) {
 void dda_step(DDA *dda) {
 
   #if ! defined ACCELERATION_TEMPORAL
-    if (dda->delta[X]) {
+    if (dda->active_axes & (1 << X)) {
       move_state.counter[X] -= dda->delta[X];
       if (move_state.counter[X] < 0) {
         x_step();
         move_state.counter[X] += dda->total_steps;
       }
 		}
-    if (dda->delta[Y]) {
+    if (dda->active_axes & (1 << Y)) {
       move_state.counter[Y] -= dda->delta[Y];
       if (move_state.counter[Y] < 0) {
         y_step();
         move_state.counter[Y] += dda->total_steps;
       }
 		}
-    if (dda->delta[Z]) {
+    if (dda->active_axes & (1 << Z)) {
       move_state.counter[Z] -= dda->delta[Z];
       if (move_state.counter[Z] < 0) {
         z_step();
         move_state.counter[Z] += dda->total_steps;
       }
 		}
-    if (dda->delta[E]) {
+    if (dda->active_axes & (1 << E)) {
       move_state.counter[E] -= dda->delta[E];
       if (move_state.counter[E] < 0) {
         e_step();
diff --git a/dda.h b/dda.h
index 0aa853f..dfa70c7 100644
--- a/dda.h
+++ b/dda.h
@@ -156,6 +156,7 @@ typedef struct {
   /// word boundaries only and fill smaller variables in between with gaps,
   /// so keep small variables grouped together to reduce the amount of these
   /// gaps. See e.g. NXP application note AN10963, page 10f.
+  uint8_t           active_axes;     ///< Bit mask of moving axes
   uint8_t           fast_axis;       ///< number of the fast axis
 
 	/// Endstop homing