aboutsummaryrefslogtreecommitdiff
path: root/plugins/gpu_unai/gpu_raster_line.h
blob: 2a7b42275c7efa8b4183b8da7e61e552e49ae1c4 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
/***************************************************************************
*   Copyright (C) 2010 PCSX4ALL Team                                      *
*   Copyright (C) 2010 Unai                                               *
*   Copyright (C) 2016 Senquack (dansilsby <AT> gmail <DOT> com)          *
*                                                                         *
*   This program is free software; you can redistribute it and/or modify  *
*   it under the terms of the GNU General Public License as published by  *
*   the Free Software Foundation; either version 2 of the License, or     *
*   (at your option) any later version.                                   *
*                                                                         *
*   This program is distributed in the hope that it will be useful,       *
*   but WITHOUT ANY WARRANTY; without even the implied warranty of        *
*   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the         *
*   GNU General Public License for more details.                          *
*                                                                         *
*   You should have received a copy of the GNU General Public License     *
*   along with this program; if not, write to the                         *
*   Free Software Foundation, Inc.,                                       *
*   51 Franklin Street, Fifth Floor, Boston, MA 02111-1307 USA.           *
***************************************************************************/

#ifndef __GPU_UNAI_GPU_RASTER_LINE_H__
#define __GPU_UNAI_GPU_RASTER_LINE_H__

///////////////////////////////////////////////////////////////////////////////
//  GPU internal line drawing functions
//
// Rewritten October 2016 by senquack:
//  Instead of one pixel at a time, lines are now drawn in runs of pixels,
//  whether vertical, horizontal, or diagonal. A new inner driver
//  'gpuPixelSpanFn' is used, as well as an enhanced Bresenham run-slice
//  algorithm. For more information, see the following:
//
//  Michael Abrash - Graphics Programming Black Book
//  Chapters 35 - 36 (does not implement diagonal runs)
//  http://www.drdobbs.com/parallel/graphics-programming-black-book/184404919
//  http://www.jagregory.com/abrash-black-book/
//
//  Article by Andrew Delong (does not implement diagonal runs)
//  http://timetraces.ca/nw/drawline.htm
//
//  'Run-Based Multi-Point Line Drawing' by Eun Jae Lee & Larry F. Hodges
//  https://smartech.gatech.edu/bitstream/handle/1853/3632/93-22.pdf
//  Provided the idea of doing a half-octant transform allowing lines with
//  slopes between 0.5 and 2.0 (diagonal runs of pixels) to be handled
//  identically to the traditional horizontal/vertical run-slice method.

// Use 16.16 fixed point precision for line math.
// NOTE: Gouraud colors used by gpuPixelSpanFn can use a different precision.
#define GPU_LINE_FIXED_BITS 16

// If defined, Gouraud lines will use fixed-point multiply-by-inverse to
// do most divisions. With enough accuracy, this should be OK.
#define USE_LINES_ALL_FIXED_PT_MATH

//////////////////////
// Flat-shaded line //
//////////////////////
void gpuDrawLineF(PtrUnion packet, const PSD gpuPixelSpanDriver)
{
	int x0, y0, x1, y1;
	int dx, dy;

	// All three of these variables should be signed (so multiplication works)
	ptrdiff_t sx;  // Sign of x delta, positive when x0 < x1
	const ptrdiff_t dst_depth  = FRAME_BYTES_PER_PIXEL; // PSX: 2 bytes per pixel
	const ptrdiff_t dst_stride = FRAME_BYTE_STRIDE;     // PSX: 2048 bytes per framebuffer line

	// Clip region: xmax/ymax seem to normally be one *past* the rightmost/
	//  bottommost pixels of the draw area. Since we render every pixel between
	//  and including both line endpoints, subtract one from xmax/ymax.
	const int xmin = gpu_unai.DrawingArea[0];
	const int ymin = gpu_unai.DrawingArea[1];
	const int xmax = gpu_unai.DrawingArea[2] - 1;
	const int ymax = gpu_unai.DrawingArea[3] - 1;

	x0 = GPU_EXPANDSIGN(packet.S2[2]) + gpu_unai.DrawingOffset[0];
	y0 = GPU_EXPANDSIGN(packet.S2[3]) + gpu_unai.DrawingOffset[1];
	x1 = GPU_EXPANDSIGN(packet.S2[4]) + gpu_unai.DrawingOffset[0];
	y1 = GPU_EXPANDSIGN(packet.S2[5]) + gpu_unai.DrawingOffset[1];

	// Always draw top to bottom, so ensure y0 <= y1
	if (y0 > y1) {
		SwapValues(y0, y1);
		SwapValues(x0, x1);
	}

	// Is line totally outside Y clipping range?
	if (y0 > ymax || y1 < ymin) return;

	dx = x1 - x0;
	dy = y1 - y0;

	// X-axis range check : max distance between any two X coords is 1023
	// (PSX hardware will not render anything violating this rule)
	// NOTE: We'll check y coord range further below
	if (dx >= CHKMAX_X || dx <= -CHKMAX_X)
		return;

	// Y-axis range check and clipping
	if (dy) {
		// Y-axis range check : max distance between any two Y coords is 511
		// (PSX hardware will not render anything violating this rule)
		if (dy >= CHKMAX_Y)
			return;

		// We already know y0 < y1
		if (y0 < ymin) {
			x0 += GPU_FAST_DIV(((ymin - y0) * dx), dy);
			y0 = ymin;
		}
		if (y1 > ymax) {
			x1 += GPU_FAST_DIV(((ymax - y1) * dx), dy);
			y1 = ymax;
		}

		// Recompute in case clipping occurred:
		dx = x1 - x0;
		dy = y1 - y0;
	}

	// Check X clipping range, set 'sx' x-direction variable
	if (dx == 0) {
		// Is vertical line totally outside X clipping range?
		if (x0 < xmin || x0 > xmax)
			return;
		sx = 0;
	} else {
		if (dx > 0) {
			// x0 is leftmost coordinate
			if (x0 > xmax) return; // Both points outside X clip range

			if (x0 < xmin) {
				if (x1 < xmin) return; // Both points outside X clip range
				y0 += GPU_FAST_DIV(((xmin - x0) * dy), dx);
				x0 = xmin;
			}

			if (x1 > xmax) {
				y1 += GPU_FAST_DIV(((xmax - x1) * dy), dx);
				x1 = xmax;
			}

			sx = +1;
			dx = x1 - x0; // Get final value, which should also be absolute value
		} else {
			// x1 is leftmost coordinate
			if (x1 > xmax) return; // Both points outside X clip range

			if (x1 < xmin) {
				if (x0 < xmin) return; // Both points outside X clip range

				y1 += GPU_FAST_DIV(((xmin - x1) * dy), dx);
				x1 = xmin;
			}

			if (x0 > xmax) {
				y0 += GPU_FAST_DIV(((xmax - x0) * dy), dx);
				x0 = xmax;
			}

			sx = -1;
			dx = x0 - x1; // Get final value, which should also be absolute value
		}

		// Recompute in case clipping occurred:
		dy = y1 - y0;
	}

	// IMPORTANT: dx,dy should now contain their absolute values

	int min_length,    // Minimum length of a pixel run
	    start_length,  // Length of first run
	    end_length,    // Length of last run
	    err_term,      // Cumulative error to determine when to draw longer run
	    err_adjup,     // Increment to err_term for each run drawn
	    err_adjdown;   // Subract this from err_term after drawing longer run

	// Color to draw with (16 bits, highest of which is unset mask bit)
	uintptr_t col16 = GPU_RGB16(packet.U4[0]);

	// We use u8 pointers even though PS1 has u16 framebuffer.
	//  This allows pixel-drawing functions to increment dst pointer
	//  directly by the passed 'incr' value, not having to shift it first.
	u8 *dst = (u8*)gpu_unai.vram + y0 * dst_stride + x0 * dst_depth;

	// SPECIAL CASE: Vertical line
	if (dx == 0) {
		gpuPixelSpanDriver(dst, col16, dst_stride, dy+1);
		return;
	}

	// SPECIAL CASE: Horizontal line
	if (dy == 0) {
		gpuPixelSpanDriver(dst, col16, sx * dst_depth, dx+1);
		return;
	}

	// SPECIAL CASE: Diagonal line
	if (dx == dy) {
		gpuPixelSpanDriver(dst, col16, dst_stride + (sx * dst_depth), dy+1);
		return;
	}

	int       major, minor;             // Major axis, minor axis
	ptrdiff_t incr_major, incr_minor;   // Ptr increment for each step along axis

	if (dx > dy) {
		major = dx;
		minor = dy;
	} else {
		major = dy;
		minor = dx;
	}

	// Determine if diagonal or horizontal runs
	if (major < (2 * minor)) {
		// Diagonal runs, so perform half-octant transformation
		minor = major - minor;

		// Advance diagonally when drawing runs
		incr_major = dst_stride + (sx * dst_depth);

		// After drawing each run, correct for over-advance along minor axis
		if (dx > dy)
			incr_minor = -dst_stride;
		else
			incr_minor = -sx * dst_depth;
	} else {
		// Horizontal or vertical runs
		if (dx > dy) {
			incr_major = sx * dst_depth;
			incr_minor = dst_stride;
		} else {
			incr_major = dst_stride;
			incr_minor = sx * dst_depth;
		}
	}

	if (minor > 1) {
		// Minimum number of pixels each run
		min_length = major / minor;

		// Initial error term; reflects an initial step of 0.5 along minor axis
		err_term = (major % minor) - (minor * 2);

		// Increment err_term this much each step along minor axis; when
		//  err_term crosses zero, draw longer pixel run.
		err_adjup = (major % minor) * 2;
	} else {
		min_length = major;
		err_term = 0;
		err_adjup = 0;
	}

	// Error term adjustment when err_term turns over; used to factor
	//  out the major-axis step made at that time
	err_adjdown = minor * 2;

	// The initial and last runs are partial, because minor axis advances
	//  only 0.5 for these runs, rather than 1. Each is half a full run,
	//  plus the initial pixel.
	start_length = end_length = (min_length / 2) + 1;

	if (min_length & 1) {
		// If there're an odd number of pixels per run, we have 1 pixel that
		//  can't be allocated to either the initial or last partial run, so
		//  we'll add 0.5 to err_term so that this pixel will be handled
		//  by the normal full-run loop
		err_term += minor;
	} else {
		// If the minimum run length is even and there's no fractional advance,
		// we have one pixel that could go to either the initial or last
		// partial run, which we arbitrarily allocate to the last run
		if (err_adjup == 0)
			start_length--; // Leave out the extra pixel at the start
	}

	// First run of pixels
	dst = gpuPixelSpanDriver(dst, col16, incr_major, start_length);
	dst += incr_minor;

	// Middle runs of pixels
	while (--minor > 0) {
		int run_length = min_length;
		err_term += err_adjup;

		// If err_term passed 0, reset it and draw longer run
		if (err_term > 0) {
			err_term -= err_adjdown;
			run_length++;
		}

		dst = gpuPixelSpanDriver(dst, col16, incr_major, run_length);
		dst += incr_minor;
	}

	// Final run of pixels
	gpuPixelSpanDriver(dst, col16, incr_major, end_length);
}

/////////////////////////
// Gouraud-shaded line //
/////////////////////////
void gpuDrawLineG(PtrUnion packet, const PSD gpuPixelSpanDriver)
{
	int x0, y0, x1, y1;
	int dx, dy, dr, dg, db;
	u32 r0, g0, b0, r1, g1, b1;

	// All three of these variables should be signed (so multiplication works)
	ptrdiff_t sx;  // Sign of x delta, positive when x0 < x1
	const ptrdiff_t dst_depth  = FRAME_BYTES_PER_PIXEL; // PSX: 2 bytes per pixel
	const ptrdiff_t dst_stride = FRAME_BYTE_STRIDE;     // PSX: 2048 bytes per framebuffer line

	// Clip region: xmax/ymax seem to normally be one *past* the rightmost/
	//  bottommost pixels of the draw area. We'll render every pixel between
	//  and including both line endpoints, so subtract one from xmax/ymax.
	const int xmin = gpu_unai.DrawingArea[0];
	const int ymin = gpu_unai.DrawingArea[1];
	const int xmax = gpu_unai.DrawingArea[2] - 1;
	const int ymax = gpu_unai.DrawingArea[3] - 1;

	x0 = GPU_EXPANDSIGN(packet.S2[2]) + gpu_unai.DrawingOffset[0];
	y0 = GPU_EXPANDSIGN(packet.S2[3]) + gpu_unai.DrawingOffset[1];
	x1 = GPU_EXPANDSIGN(packet.S2[6]) + gpu_unai.DrawingOffset[0];
	y1 = GPU_EXPANDSIGN(packet.S2[7]) + gpu_unai.DrawingOffset[1];

	u32 col0 = packet.U4[0];
	u32 col1 = packet.U4[2];

	// Always draw top to bottom, so ensure y0 <= y1
	if (y0 > y1) {
		SwapValues(y0, y1);
		SwapValues(x0, x1);
		SwapValues(col0, col1);
	}

	// Is line totally outside Y clipping range?
	if (y0 > ymax || y1 < ymin) return;

	// If defined, Gouraud colors are fixed-point 5.11, otherwise they are 8.16
	// (This is only beneficial if using SIMD-optimized pixel driver)
#ifdef GPU_GOURAUD_LOW_PRECISION
	r0 = (col0 >> 3) & 0x1f;  g0 = (col0 >> 11) & 0x1f;  b0 = (col0 >> 19) & 0x1f;
	r1 = (col1 >> 3) & 0x1f;  g1 = (col1 >> 11) & 0x1f;  b1 = (col1 >> 19) & 0x1f;
#else
	r0 = col0 & 0xff;  g0 = (col0 >> 8) & 0xff;  b0 = (col0 >> 16) & 0xff;
	r1 = col1 & 0xff;  g1 = (col1 >> 8) & 0xff;  b1 = (col1 >> 16) & 0xff;
#endif

	dx = x1 - x0;
	dy = y1 - y0;
	dr = r1 - r0;
	dg = g1 - g0;
	db = b1 - b0;

	// X-axis range check : max distance between any two X coords is 1023
	// (PSX hardware will not render anything violating this rule)
	// NOTE: We'll check y coord range further below
	if (dx >= CHKMAX_X || dx <= -CHKMAX_X)
		return;

	// Y-axis range check and clipping
	if (dy) {
		// Y-axis range check : max distance between any two Y coords is 511
		// (PSX hardware will not render anything violating this rule)
		if (dy >= CHKMAX_Y)
			return;

		// We already know y0 < y1
		if (y0 < ymin) {
#ifdef USE_LINES_ALL_FIXED_PT_MATH
			s32 factor = GPU_FAST_DIV(((ymin - y0) << GPU_LINE_FIXED_BITS), dy);
			x0 += (dx * factor) >> GPU_LINE_FIXED_BITS;
			r0 += (dr * factor) >> GPU_LINE_FIXED_BITS;
			g0 += (dg * factor) >> GPU_LINE_FIXED_BITS;
			b0 += (db * factor) >> GPU_LINE_FIXED_BITS;
#else
			x0 += (ymin - y0) * dx / dy;
			r0 += (ymin - y0) * dr / dy;
			g0 += (ymin - y0) * dg / dy;
			b0 += (ymin - y0) * db / dy;
#endif
			y0 = ymin;
		}

		if (y1 > ymax) {
#ifdef USE_LINES_ALL_FIXED_PT_MATH
			s32 factor = GPU_FAST_DIV(((ymax - y1) << GPU_LINE_FIXED_BITS), dy);
			x1 += (dx * factor) >> GPU_LINE_FIXED_BITS;
			r1 += (dr * factor) >> GPU_LINE_FIXED_BITS;
			g1 += (dg * factor) >> GPU_LINE_FIXED_BITS;
			b1 += (db * factor) >> GPU_LINE_FIXED_BITS;
#else
			x1 += (ymax - y1) * dx / dy;
			r1 += (ymax - y1) * dr / dy;
			g1 += (ymax - y1) * dg / dy;
			b1 += (ymax - y1) * db / dy;
#endif
			y1 = ymax;
		}

		// Recompute in case clipping occurred:
		dx = x1 - x0;
		dy = y1 - y0;
		dr = r1 - r0;
		dg = g1 - g0;
		db = b1 - b0;
	}

	// Check X clipping range, set 'sx' x-direction variable
	if (dx == 0) {
		// Is vertical line totally outside X clipping range?
		if (x0 < xmin || x0 > xmax)
			return;
		sx = 0;
	} else {
		if (dx > 0) {
			// x0 is leftmost coordinate
			if (x0 > xmax) return; // Both points outside X clip range

			if (x0 < xmin) {
				if (x1 < xmin) return; // Both points outside X clip range

#ifdef USE_LINES_ALL_FIXED_PT_MATH
				s32 factor = GPU_FAST_DIV(((xmin - x0) << GPU_LINE_FIXED_BITS), dx);
				y0 += (dy * factor) >> GPU_LINE_FIXED_BITS;
				r0 += (dr * factor) >> GPU_LINE_FIXED_BITS;
				g0 += (dg * factor) >> GPU_LINE_FIXED_BITS;
				b0 += (db * factor) >> GPU_LINE_FIXED_BITS;
#else
				y0 += (xmin - x0) * dy / dx;
				r0 += (xmin - x0) * dr / dx;
				g0 += (xmin - x0) * dg / dx;
				b0 += (xmin - x0) * db / dx;
#endif
				x0 = xmin;
			}

			if (x1 > xmax) {
#ifdef USE_LINES_ALL_FIXED_PT_MATH
				s32 factor = GPU_FAST_DIV(((xmax - x1) << GPU_LINE_FIXED_BITS), dx);
				y1 += (dy * factor) >> GPU_LINE_FIXED_BITS;
				r1 += (dr * factor) >> GPU_LINE_FIXED_BITS;
				g1 += (dg * factor) >> GPU_LINE_FIXED_BITS;
				b1 += (db * factor) >> GPU_LINE_FIXED_BITS;
#else
				y1 += (xmax - x1) * dy / dx;
				r1 += (xmax - x1) * dr / dx;
				g1 += (xmax - x1) * dg / dx;
				b1 += (xmax - x1) * db / dx;
#endif
				x1 = xmax;
			}

			sx = +1;
			dx = x1 - x0; // Get final value, which should also be absolute value
		} else {
			// x1 is leftmost coordinate
			if (x1 > xmax) return; // Both points outside X clip range

			if (x1 < xmin) {
				if (x0 < xmin) return; // Both points outside X clip range

#ifdef USE_LINES_ALL_FIXED_PT_MATH
				s32 factor = GPU_FAST_DIV(((xmin - x1) << GPU_LINE_FIXED_BITS), dx);
				y1 += (dy * factor) >> GPU_LINE_FIXED_BITS;
				r1 += (dr * factor) >> GPU_LINE_FIXED_BITS;
				g1 += (dg * factor) >> GPU_LINE_FIXED_BITS;
				b1 += (db * factor) >> GPU_LINE_FIXED_BITS;
#else
				y1 += (xmin - x1) * dy / dx;
				r1 += (xmin - x1) * dr / dx;
				g1 += (xmin - x1) * dg / dx;
				b1 += (xmin - x1) * db / dx;
#endif
				x1 = xmin;
			}

			if (x0 > xmax) {
#ifdef USE_LINES_ALL_FIXED_PT_MATH
				s32 factor = GPU_FAST_DIV(((xmax - x0) << GPU_LINE_FIXED_BITS), dx);
				y0 += (dy * factor) >> GPU_LINE_FIXED_BITS;
				r0 += (dr * factor) >> GPU_LINE_FIXED_BITS;
				g0 += (dg * factor) >> GPU_LINE_FIXED_BITS;
				b0 += (db * factor) >> GPU_LINE_FIXED_BITS;
#else
				y0 += (xmax - x0) * dy / dx;
				r0 += (xmax - x0) * dr / dx;
				g0 += (xmax - x0) * dg / dx;
				b0 += (xmax - x0) * db / dx;
#endif
				x0 = xmax;
			}

			sx = -1;
			dx = x0 - x1; // Get final value, which should also be absolute value
		}

		// Recompute in case clipping occurred:
		dy = y1 - y0;
		dr = r1 - r0;
		dg = g1 - g0;
		db = b1 - b0;
	}

	// IMPORTANT: dx,dy should now contain their absolute values

	int min_length,    // Minimum length of a pixel run
	    start_length,  // Length of first run
	    end_length,    // Length of last run
	    err_term,      // Cumulative error to determine when to draw longer run
	    err_adjup,     // Increment to err_term for each run drawn
	    err_adjdown;   // Subract this from err_term after drawing longer run

	GouraudColor gcol;
	gcol.r = r0 << GPU_GOURAUD_FIXED_BITS;
	gcol.g = g0 << GPU_GOURAUD_FIXED_BITS;
	gcol.b = b0 << GPU_GOURAUD_FIXED_BITS;

	// We use u8 pointers even though PS1 has u16 framebuffer.
	//  This allows pixel-drawing functions to increment dst pointer
	//  directly by the passed 'incr' value, not having to shift it first.
	u8 *dst = (u8*)gpu_unai.vram + y0 * dst_stride + x0 * dst_depth;

	// SPECIAL CASE: Vertical line
	if (dx == 0) {
#ifdef USE_LINES_ALL_FIXED_PT_MATH
		// Get dy fixed-point inverse
		s32 inv_factor = 1 << GPU_GOURAUD_FIXED_BITS;
		if (dy > 1) inv_factor = GPU_FAST_DIV(inv_factor, dy);

		// Simultaneously divide and convert integer to Gouraud fixed point:
		gcol.r_incr = dr * inv_factor;
		gcol.g_incr = dg * inv_factor;
		gcol.b_incr = db * inv_factor;
#else
		// First, convert to Gouraud fixed point
		gcol.r_incr = dr << GPU_GOURAUD_FIXED_BITS;
		gcol.g_incr = dg << GPU_GOURAUD_FIXED_BITS;
		gcol.b_incr = db << GPU_GOURAUD_FIXED_BITS;

		if (dy > 1) {
			if (dr) gcol.r_incr /= dy;
			if (dg) gcol.g_incr /= dy;
			if (db) gcol.b_incr /= dy;
		}
#endif
		
		gpuPixelSpanDriver(dst, (uintptr_t)&gcol, dst_stride, dy+1);
		return;
	}

	// SPECIAL CASE: Horizontal line
	if (dy == 0) {
#ifdef USE_LINES_ALL_FIXED_PT_MATH
		// Get dx fixed-point inverse
		s32 inv_factor = (1 << GPU_GOURAUD_FIXED_BITS);
		if (dx > 1) inv_factor = GPU_FAST_DIV(inv_factor, dx);

		// Simultaneously divide and convert integer to Gouraud fixed point:
		gcol.r_incr = dr * inv_factor;
		gcol.g_incr = dg * inv_factor;
		gcol.b_incr = db * inv_factor;
#else
		gcol.r_incr = dr << GPU_GOURAUD_FIXED_BITS;
		gcol.g_incr = dg << GPU_GOURAUD_FIXED_BITS;
		gcol.b_incr = db << GPU_GOURAUD_FIXED_BITS;

		if (dx > 1) {
			if (dr) gcol.r_incr /= dx;
			if (dg) gcol.g_incr /= dx;
			if (db) gcol.b_incr /= dx;
		}
#endif

		gpuPixelSpanDriver(dst, (uintptr_t)&gcol, sx * dst_depth, dx+1);
		return;
	}

	// SPECIAL CASE: Diagonal line
	if (dx == dy) {
#ifdef USE_LINES_ALL_FIXED_PT_MATH
		// Get dx fixed-point inverse
		s32 inv_factor = (1 << GPU_GOURAUD_FIXED_BITS);
		if (dx > 1) inv_factor = GPU_FAST_DIV(inv_factor, dx);

		// Simultaneously divide and convert integer to Gouraud fixed point:
		gcol.r_incr = dr * inv_factor;
		gcol.g_incr = dg * inv_factor;
		gcol.b_incr = db * inv_factor;
#else
		// First, convert to Gouraud fixed point
		gcol.r_incr = dr << GPU_GOURAUD_FIXED_BITS;
		gcol.g_incr = dg << GPU_GOURAUD_FIXED_BITS;
		gcol.b_incr = db << GPU_GOURAUD_FIXED_BITS;

		if (dx > 1) {
			if (dr) gcol.r_incr /= dx;
			if (dg) gcol.g_incr /= dx;
			if (db) gcol.b_incr /= dx;
		}
#endif

		gpuPixelSpanDriver(dst, (uintptr_t)&gcol, dst_stride + (sx * dst_depth), dy+1);
		return;
	}

	int       major, minor;             // Absolute val of major,minor axis delta
	ptrdiff_t incr_major, incr_minor;   // Ptr increment for each step along axis

	if (dx > dy) {
		major = dx;
		minor = dy;
	} else {
		major = dy;
		minor = dx;
	}

	// Determine if diagonal or horizontal runs
	if (major < (2 * minor)) {
		// Diagonal runs, so perform half-octant transformation
		minor = major - minor;

		// Advance diagonally when drawing runs
		incr_major = dst_stride + (sx * dst_depth);

		// After drawing each run, correct for over-advance along minor axis
		if (dx > dy)
			incr_minor = -dst_stride;
		else
			incr_minor = -sx * dst_depth;
	} else {
		// Horizontal or vertical runs
		if (dx > dy) {
			incr_major = sx * dst_depth;
			incr_minor = dst_stride;
		} else {
			incr_major = dst_stride;
			incr_minor = sx * dst_depth;
		}
	}

#ifdef USE_LINES_ALL_FIXED_PT_MATH
	s32 major_inv = GPU_FAST_DIV((1 << GPU_GOURAUD_FIXED_BITS), major);

	// Simultaneously divide and convert from integer to Gouraud fixed point:
	gcol.r_incr = dr * major_inv;
	gcol.g_incr = dg * major_inv;
	gcol.b_incr = db * major_inv;
#else
	gcol.r_incr = dr ? ((dr << GPU_GOURAUD_FIXED_BITS) / major) : 0;
	gcol.g_incr = dg ? ((dg << GPU_GOURAUD_FIXED_BITS) / major) : 0;
	gcol.b_incr = db ? ((db << GPU_GOURAUD_FIXED_BITS) / major) : 0;
#endif

	if (minor > 1) {
		// Minimum number of pixels each run
		min_length = major / minor;

		// Initial error term; reflects an initial step of 0.5 along minor axis
		err_term = (major % minor) - (minor * 2);

		// Increment err_term this much each step along minor axis; when
		//  err_term crosses zero, draw longer pixel run.
		err_adjup = (major % minor) * 2;
	} else {
		min_length = major;
		err_term = 0;
		err_adjup = 0;
	}

	// Error term adjustment when err_term turns over; used to factor
	//  out the major-axis step made at that time
	err_adjdown = minor * 2;

	// The initial and last runs are partial, because minor axis advances
	//  only 0.5 for these runs, rather than 1. Each is half a full run,
	//  plus the initial pixel.
	start_length = end_length = (min_length / 2) + 1;

	if (min_length & 1) {
		// If there're an odd number of pixels per run, we have 1 pixel that
		//  can't be allocated to either the initial or last partial run, so
		//  we'll add 0.5 to err_term so that this pixel will be handled
		//  by the normal full-run loop
		err_term += minor;
	} else {
		// If the minimum run length is even and there's no fractional advance,
		// we have one pixel that could go to either the initial or last
		// partial run, which we'll arbitrarily allocate to the last run
		if (err_adjup == 0)
			start_length--; // Leave out the extra pixel at the start
	}

	// First run of pixels
	dst = gpuPixelSpanDriver(dst, (uintptr_t)&gcol, incr_major, start_length);
	dst += incr_minor;

	// Middle runs of pixels
	while (--minor > 0) {
		int run_length = min_length;
		err_term += err_adjup;

		// If err_term passed 0, reset it and draw longer run
		if (err_term > 0) {
			err_term -= err_adjdown;
			run_length++;
		}

		dst = gpuPixelSpanDriver(dst, (uintptr_t)&gcol, incr_major, run_length);
		dst += incr_minor;
	}

	// Final run of pixels
	gpuPixelSpanDriver(dst, (uintptr_t)&gcol, incr_major, end_length);
}

#endif /* __GPU_UNAI_GPU_RASTER_LINE_H__ */