From: John Tsiombikas <nuclear@member.fsf.org>
Date: Thu, 22 Apr 2021 09:01:25 +0000 (+0300)
Subject: Merge branch 'tunsweep'
X-Git-Url: http://git.mutantstargoat.com/user/nuclear/?a=commitdiff_plain;h=e8d7821317677f1d0fd6ee9efa9b9ef0ef006d01;hp=ff74cf463fa44906add9344891096cada0970205;p=gbajam21

Merge branch 'tunsweep'
---

diff --git a/Makefile b/Makefile
index 9da7f4d..68b1f63 100644
--- a/Makefile
+++ b/Makefile
@@ -55,7 +55,7 @@ tools/lutgen: tools/lutgen.c
 	cc -o $@ $< -lm
 
 tools/tungen: tools/tungen.c
-	cc -o $@ $< -lm
+	cc -o $@ -O3 -fopenmp $< -lm -lgomp -lpthread
 
 tools/mmutil/mmutil:
 	$(MAKE) -C tools/mmutil
@@ -73,7 +73,7 @@ data/lut.s: tools/lutgen
 	tools/lutgen >$@
 
 data/tun.map: tools/tungen
-	tools/tungen -s 256x256 >$@
+	tools/tungen -s 240x160 -y -c 400 -n 32 >$@
 
 data/snd.bin: $(audata) tools/mmutil/mmutil
 	tools/mmutil/mmutil -o$@ -hdata/snd.h $(audata)
diff --git a/src/gamescr.c b/src/gamescr.c
index 9e9d6e1..2744d5d 100644
--- a/src/gamescr.c
+++ b/src/gamescr.c
@@ -93,66 +93,83 @@ void gamescr(void)
 __attribute__((noinline, target("arm"), section(".iwram")))
 static void draw_tunnel(void)
 {
-	int i, j, tx, ty, angle, depth, zoffs, uoffs;
-	uint16_t pptop;
-	uint16_t *top;
-	uint16_t tun;
-	int32_t startx, starty;
-	uint16_t *tmap = (uint16_t*)tunmap;
+	int i, j, tx, ty, angle, depth, zoffs, uoffs, flip, tunturn;
+	static int tunsweep;
+	uint16_t pptop, ppbot;
+	uint16_t *top, *bot;
+	uint32_t tun, *tunptr;
 
-	zoffs = nframes;
+	//tunsweep = SIN(nframes) >> 4;
+
+	if((bnstate & BN_RT) && tunsweep > -31) tunsweep--;
+	if((bnstate & BN_LT) && tunsweep < 31) tunsweep++;
 
-	/*
-	if(bnstate & KEY_LT) tunrot++;
-	if(bnstate & KEY_RT) tunrot--;
-	*/
-	tunrot = nframes;
+	flip = tunsweep < 0;
+	tunturn = abs(tunsweep) & 0x1f;
 
-	tunmat[0] = COS(tunrot);
-	tunmat[1] = -SIN(tunrot);
-	tunmat[2] = SIN(tunrot);
-	tunmat[3] = COS(tunrot);
-	tunx = (128 << 8) - tunmat[0] * 120 + tunmat[1] * -80;
-	tuny = (128 << 8) - tunmat[2] * 120 + tunmat[3] * -80;
+	zoffs = nframes;
 
-	uoffs = tunrot;
+	uoffs = flip ? -nframes : nframes;
 
 	top = vram[backbuf];
-	for(i=0; i<160; i++) {
-		startx = tunx;
-		starty = tuny;
-		for(j=0; j<240/2; j++) {
-			tx = (tunx >> 8) & 0xff;
-			ty = (tuny >> 8) & 0xff;
-			tun = tmap[(ty << 8) + tx];
-
-			tunx += tunmat[0];
-			tuny += tunmat[2];
-
-			angle = tun & 0xff;
-			depth = (tun >> 8) & 0xff;
-			tx = ((angle >> 1) - uoffs) & 0x1f;
-			ty = ((depth >> 1) + zoffs) & 0x1f;
-			pptop = tex[(ty << 5) + tx];
-
-			tx = (tunx >> 8) & 0xff;
-			ty = (tuny >> 8) & 0xff;
-			tun = tmap[(ty << 8) + tx];
-
-			tunx += tunmat[0];
-			tuny += tunmat[2];
-
-			angle = tun & 0xff;
-			depth = (tun >> 8) & 0xff;
-			tx = ((angle >> 1) - uoffs) & 0x1f;
-			ty = ((depth >> 1) + zoffs) & 0x1f;
-			pptop |= (uint16_t)tex[(ty << 5) + tx] << 8;
-
-			*top++ = pptop;
+	bot = vram[backbuf] + 159 * 240 / 2;
+	tunptr = tunmap + tunturn * 9600;
+
+	if(flip) {
+		tunptr += 240/2;
+		for(i=0; i<80; i++) {
+			for(j=0; j<240/2; j++) {
+				tun = *--tunptr;
+				tun = (tun >> 16) | (tun << 16);
+
+				angle = tun & 0xff;
+				depth = (tun >> 8) & 0xff;
+				tx = ~((angle >> 1) - uoffs) & 0x1f;
+				ty = ((depth >> 1) + zoffs) & 0x1f;
+				pptop = tex[(ty << 5) + tx];
+				tx = ((angle >> 1) + uoffs) & 0x1f;
+				ppbot = tex[(ty << 5) + tx];
+
+				angle = (tun >> 16) & 0xff;
+				depth = (tun >> 24) & 0xff;
+				tx = ~((angle >> 1) - uoffs) & 0x1f;
+				ty = ((depth >> 1) + zoffs) & 0x1f;
+				pptop |= (uint16_t)tex[(ty << 5) + tx] << 8;
+				tx = ((angle >> 1) + uoffs) & 0x1f;
+				ppbot |= (uint16_t)tex[(ty << 5) + tx] << 8;
+
+				*top++ = pptop;
+				*bot++ = ppbot;
+			}
+			bot -= 240;
+			tunptr += 240;
+		}
+	} else {
+		for(i=0; i<80; i++) {
+			for(j=0; j<240/2; j++) {
+				tun = *tunptr++;
+
+				angle = tun & 0xff;
+				depth = (tun >> 8) & 0xff;
+				tx = ((angle >> 1) - uoffs) & 0x1f;
+				ty = ((depth >> 1) + zoffs) & 0x1f;
+				pptop = tex[(ty << 5) + tx];
+				tx = ~((angle >> 1) + uoffs) & 0x1f;
+				ppbot = tex[(ty << 5) + tx];
+
+				angle = (tun >> 16) & 0xff;
+				depth = (tun >> 24) & 0xff;
+				tx = ((angle >> 1) - uoffs) & 0x1f;
+				ty = ((depth >> 1) + zoffs) & 0x1f;
+				pptop |= (uint16_t)tex[(ty << 5) + tx] << 8;
+				tx = ~((angle >> 1) + uoffs) & 0x1f;
+				ppbot |= (uint16_t)tex[(ty << 5) + tx] << 8;
+
+				*top++ = pptop;
+				*bot++ = ppbot;
+			}
+			bot -= 240;
 		}
-
-		tunx = startx + tunmat[1];
-		tuny = starty + tunmat[3];
 	}
 }
 
diff --git a/tools/tungen.c b/tools/tungen.c
index 62e0d63..f480758 100644
--- a/tools/tungen.c
+++ b/tools/tungen.c
@@ -9,10 +9,14 @@ struct vec2 {
 
 int main(int argc, char **argv)
 {
-	int i, j, imgrad, xsz = 240, ysz = 160, texsz = 128;
+	int i, j, frm, imgrad, out_nlines, xsz = 240, ysz = 160, texsz = 128;
+	int half_y = 0;
+	int center = 0;
 	struct vec2 *tunbuf, *tun;
-	float aspect, prev_r;
+	float prev_r;
 	struct vec2 *buf, *ptr;
+	char *endp;
+	int num_frames = 1;
 
 	for(i=1; i<argc; i++) {
 		if(argv[i][0] == '-') {
@@ -32,6 +36,33 @@ int main(int argc, char **argv)
 				}
 				break;
 
+			case 'y':
+				half_y = 1;
+				break;
+
+			case 'c':
+				if(!argv[++i]) {
+					fprintf(stderr, "-c must be followed by a center pixel\n");
+					return 1;
+				}
+				center = strtol(argv[i], &endp, 10);
+				if(endp == argv[i]) {
+					fprintf(stderr, "-c invalid center position: %s\n", argv[i]);
+					return 1;
+				}
+				break;
+
+			case 'n':
+				if(!argv[++i]) {
+					fprintf(stderr, "-n must be followed by the number of frames\n");
+					return 1;
+				}
+				if(!(num_frames = atoi(argv[i]))) {
+					fprintf(stderr, "-n invalid number of frames: %s\n", argv[i]);
+					return 1;
+				}
+				break;
+
 			default:
 				goto invalopt;
 			}
@@ -41,71 +72,78 @@ invalopt:	fprintf(stderr, "invalid argument: %s\n", argv[i]);
 		}
 	}
 
+	out_nlines = half_y ? ysz / 2 : ysz;
+
 	if(!(buf = malloc(xsz * ysz * sizeof *buf))) {
 		perror("failed to allocate buffer");
 		return 1;
 	}
 	imgrad = sqrt(xsz * xsz + ysz * ysz);
 
-#define UDIV	2048
-#define VDIV	32768
-	prev_r = 0.0f;
-	for(i=0; i<VDIV; i++) {
-		float v = (float)(VDIV - i) / (float)VDIV;
-		float r = 4.0 / v + 16;
-		float z = v * 400.0f;
-
-		/* don't bother drawing rings < 1 pixel apart */
-		if(fabs(r - prev_r) < 0.05) continue;
-
-		for(j=0; j<UDIV; j++) {
-			float u = (float)j / (float)(UDIV - 1);
-			float theta = 2.0f * u * M_PI;
-
-			int x = (int)(cos(theta) * r - z) + xsz / 2;
-			int y = (int)(sin(theta) * r) + ysz / 2;
-
-			if(x >= 0 && x < xsz && y >= 0 && y < ysz) {
-				ptr = buf + y * xsz + x;
-				ptr->x = u;
-				ptr->y = v * 8;
-			}
-		}
-		prev_r = r;
-	}
-
 	FILE *fp = fopen("tun_preview.ppm", "wb");
 	if(fp) {
-		fprintf(fp, "P6\n%d %d\n255\n", xsz, ysz);
+		fprintf(fp, "P6\n%d %d\n255\n", xsz, out_nlines * num_frames);
 	}
 
-	aspect = (float)xsz / (float)ysz;
-
-	ptr = buf;
-	for(i=0; i<ysz; i++) {
-		for(j=0; j<xsz; j++) {
-			float u = ptr->x;
-			float v = ptr->y;
-			int r = (int)(u * 8.0 * 255.0f) & 0xff;
-			int g = (int)(v * 8.0 * 255.0f) & 0xff;
-			int b = (~(int)(v * 0.5 * 255.0f) & 0xff) + 105;
-			if(b > 255) b = 255;
-			if(b < 0) b = 0;
-
-			/*if(v > 2.0) r = g = b = 0;*/
 
-			ptr++;
+	for(frm=0; frm<num_frames; frm++) {
+		int coffs = num_frames > 1 ? frm * center / (num_frames - 1) : center;
 
-			uint16_t out = ((uint16_t)(u * 255.0f) & 0xff) |
-				(((uint16_t)(v * 255.0f) & 0xff) << 8);
-			fwrite(&out, sizeof out, 1, stdout);
+#define UDIV	2048
+#define VDIV	32768
+		prev_r = 0.0f;
+#pragma omp parallel for private(i, j, prev_r, ptr) schedule(dynamic)
+		for(i=0; i<VDIV; i++) {
+			float v = (float)(VDIV - i) / (float)VDIV;
+			float r = 4.0 / v + 16;
+			float z = v * coffs;
+
+			/* don't bother drawing rings < 1 pixel apart */
+			if(fabs(r - prev_r) < 0.05) continue;
+
+			for(j=0; j<UDIV; j++) {
+				float u = (float)j / (float)UDIV;
+				float theta = 2.0f * u * M_PI;
+
+				int x = (int)(cos(theta) * r - z) + xsz / 2;
+				int y = (int)(sin(theta) * r) + ysz / 2;
+
+				if(x >= 0 && x < xsz && y >= 0 && y < ysz) {
+					ptr = buf + y * xsz + x;
+					ptr->x = u;
+					ptr->y = v * 8;
+				}
+			}
+			prev_r = r;
+		}
 
-			if(fp) {
-				fputc(r, fp);
-				fputc(g, fp);
-				fputc(b, fp);
+		ptr = buf;
+		for(i=0; i<out_nlines; i++) {
+			for(j=0; j<xsz; j++) {
+				float u = ptr->x;
+				float v = ptr->y;
+				int r = (int)(u * 8.0 * 255.0f) & 0xff;
+				int g = (int)(v * 8.0 * 255.0f) & 0xff;
+				int b = (~(int)(v * 0.5 * 255.0f) & 0xff) + 105;
+				if(b > 255) b = 255;
+				if(b < 0) b = 0;
+
+				/*if(v > 2.0) r = g = b = 0;*/
+
+				ptr++;
+
+				uint16_t out = ((uint16_t)(u * 255.0f) & 0xff) |
+					(((uint16_t)(v * 255.0f) & 0xff) << 8);
+				fwrite(&out, sizeof out, 1, stdout);
+
+				if(fp) {
+					fputc(r, fp);
+					fputc(g, fp);
+					fputc(b, fp);
+				}
 			}
 		}
+
 	}
 	fflush(stdout);