initial commit

2014-01-18 15:06:11 +01:00
parent 3fabb8dcf8
commit 768bec39b3
408 changed files with 171325 additions and 2 deletions
--- a/demo/src9/Makefile
+++ b/demo/src9/Makefile
@@ -0,0 +1,38 @@
+export PATH := $(DEVKITARM)/bin:$(PATH)
+
+AS = arm-eabi-gcc
+CC = arm-eabi-gcc
+LD = arm-eabi-ld
+OBJCOPY = arm-eabi-objcopy
+
+ARCH = -mthumb -mthumb-interwork
+CFLAGS = -Wall -Os -DARM9 -march=armv5te -mtune=arm946e-s
+CFLAGS += -nostdlib -nodefaultlibs -nostartfiles -fomit-frame-pointer
+INCLUDE = -Isrc -I$(DEVKITPRO)/libnds/include
+ASFLAGS = -x assembler-with-cpp -I src
+LDFLAGS = -mno-fpu -specs=ds_arm9.specs
+LDFLAGS += -nostdlib -nodefaultlibs -nostartfiles
+LDFLAGS += -Wl,-Map,arm9.map
+
+ASMFILES = $(wildcard src/*.s)
+CFILES = $(wildcard src/*.c)
+OBJS = $(CFILES:.c=.o) $(ASMFILES:.s=.o)
+
+#TODO : add auto asm dump
+.c.o: %.c
+	@echo Compiling $<
+	@$(CC) $(ARCH) $(CFLAGS) $(INCLUDE) -o $@ -c $<
+
+.s.o: %.s
+	@echo Assembling $<
+	@$(AS) $(ASFLAGS) -o $@ -c $<
+
+bin: elf
+	@$(OBJCOPY) -O binary $< $@
+
+elf: $(OBJS)
+	@echo -e '\e[1;33mLinking...\e[0m'
+	@$(CC) $(ARCH) $(LDFLAGS) -o $@ $(OBJS)
+
+clean:
+	rm -f elf bin $(OBJS)
--- a/demo/src9/src/3D.h
+++ b/demo/src9/src/3D.h
@@ -0,0 +1,73 @@
+#ifndef _3D_H_
+#define _3D_H_
+
+#include <nds.h>
+
+// Commands
+#define CMTXMODE 		0x10
+#define CMTXPUSH 		0x11
+#define CMTXPOP			0x12
+#define CMTXIDENTITY 	0x15
+#define CMTXLOAD4x4 	0x16
+#define CMTXLOAD4x3 	0x17
+#define CMTXMULT3x3 	0x1A
+#define CMTXSCALE 		0x1B
+#define CMTXTRANSLATE 	0x1C
+#define CCOLOR			0x20
+#define CNORMAL 		0x21
+#define CVERTEX 		0x24
+#define CPOLYATTR 		0x29
+#define CDIFAMB 		0x30
+#define CLIGHTVECTOR	0x32
+#define CLIGHTCOLOR 	0x33
+#define CBEGIN			0x40
+#define CVIEWPORT 		0x60
+
+// Parameters
+#define PROJECTION 	0
+#define MODELVIEW 	2
+
+#define QUAD 1
+
+#define NORMAL(x,y,z) (((x)&0x3FF)|(((y)&0x3FF)<<10)|(((z)&0x3FF)<<20)) 
+#define VERTEX(x,y,z) \
+	(((x)&0x3FF)|(((y)&0x3FF)<<10)|(((z)&0x3FF)<<20))
+
+#define DIFAMB(dif,amb) ((dif)|((amb)<<16))
+
+#define VIEWPORT(x1,y1,x2,y2) ((x1)|((y1)<<8)|((x2)<<16)|((y2)<<24))
+
+// Polygon attributes
+#define LIGHT0 BIT(0)
+#define POLYFRONT BIT(7)
+#define POLYBACK BIT(6)
+#define SOLID (31<<16)
+#define ALPHA(a) ((a)<<16)
+#define WIREFRAME (0<<16)
+
+// GX FIFO command packer
+#define COMMAND(a,b,c,d) \
+	(((a)&0xFF)|(((b)&0xFF)<<8)|(((c)&0xFF)<<16)|(((d)&0xFF)<<24))
+
+void call_list(u32 * list, int size);
+
+#define push() MATRIX_PUSH = 0
+#define pop() MATRIX_POP = 1
+#define identity() MATRIX_IDENTITY = 0
+#define swap() GFX_FLUSH = 0
+#define translate(x, y, z) \
+	MATRIX_TRANSLATE = x; \
+	MATRIX_TRANSLATE = y; \
+	MATRIX_TRANSLATE = z
+#define scale(x, y, z) \
+	MATRIX_SCALE = x; \
+	MATRIX_SCALE = y; \
+	MATRIX_SCALE = z
+
+void rotate(int angle, int v);
+
+#define rotateX(angle) rotate(angle, 1)
+#define rotateY(angle) rotate(angle, 2)
+#define rotateZ(angle) rotate(angle, 3)
+
+#endif // _3D_H_
--- a/demo/src9/src/3D.s
+++ b/demo/src9/src/3D.s
@@ -0,0 +1,25 @@
+	.arch armv5te
+	.thumb
+	.text
+	.align 2
+	.global call_list
+	.thumb_func
+	.type call_list, %function
+call_list:
+	ldr	r3, dmasrc
+	str	r0, [r3]
+	ldr	r2, fifo
+	str	r2, [r3, #4]
+	ldr	r2, param
+	orr	r1, r2
+	str	r1, [r3, #8]
+busy:
+	ldr	r2, [r3, #8]
+	cmp	r2, #0
+	blt	busy
+	bx lr
+.L6:
+	.align	2
+dmasrc:	.word 0x40000B0
+fifo:	.word 0x4000400
+param:	.word (1<<31)|(7<<27)|(1<<22)|(1<<26) @; enable, startfifo, destfix, 32bit
--- a/demo/src9/src/ds_arm9_crt0.s
+++ b/demo/src9/src/ds_arm9_crt0.s
@@ -0,0 +1,203 @@
+#include "options.h"
+@--------------------------------------------------------------------------------
+@ DS processor selection
+@--------------------------------------------------------------------------------
+	.arch	armv5te
+	.cpu	arm946e-s
+@--------------------------------------------------------------------------------
+#define PAGE_4K		(0b01011 << 1)
+#define PAGE_8K		(0b01100 << 1)
+#define PAGE_16K	(0b01101 << 1)
+#define PAGE_32K	(0b01110 << 1)
+#define PAGE_64K	(0b00111 << 1)
+#define PAGE_128K	(0b10000 << 1)
+#define PAGE_256K	(0b10001 << 1)
+#define PAGE_512K	(0b10010 << 1)
+#define PAGE_1M		(0b10011 << 1)
+#define PAGE_2M		(0b10100 << 1)
+#define PAGE_4M		(0b10101 << 1)
+#define PAGE_8M		(0b10110 << 1)
+#define PAGE_16M	(0b10111 << 1)
+#define PAGE_32M	(0b11000 << 1)
+#define PAGE_64M	(0b11001 << 1)
+#define PAGE_128M	(0b11010 << 1)
+#define PAGE_256M	(0b11011 << 1)
+#define PAGE_512M	(0b11100 << 1)
+#define PAGE_1G		(0b11101 << 1)
+#define PAGE_2G		(0b11110 << 1)
+#define PAGE_4G		(0b11111 << 1)
+
+#define ITCM_LOAD		(1<<19)
+#define ITCM_ENABLE		(1<<18)
+#define DTCM_LOAD		(1<<17)
+#define DTCM_ENABLE		(1<<16)
+#define DISABLE_TBIT	(1<<15)
+#define ROUND_ROBIN		(1<<14)
+#define ALT_VECTORS		(1<<13)
+#define ICACHE_ENABLE	(1<<12)
+#define BIG_ENDIAN		(1<<7)
+#define DCACHE_ENABLE	(1<<2)
+#define PROTECT_ENABLE	(1<<0)
+
+#if SAFE
+	.equ	_libnds_argv,0x027FFF70
+#endif
+
+@--------------------------------------------------------------------------------
+	.section ".init"
+	.global _start
+@--------------------------------------------------------------------------------
+	.align	4
+	.arm
+@--------------------------------------------------------------------------------
+_start:
+@--------------------------------------------------------------------------------
+#if SAFE
+	mov	r0, #0x04000000			@ IME = 0;
+	str	r0, [r0, #0x208]
+
+@--------------------------------------------------------------------------------
+@ turn the power on for M3
+@--------------------------------------------------------------------------------
+	ldr     r1, =0x8203
+	add	r0,r0,#0x304
+	strh    r1, [r0]
+
+	ldr	r1, =0x00002078			@ disable TCM and protection unit
+	mcr	p15, 0, r1, c1, c0
+
+@--------------------------------------------------------------------------------
+@ Protection Unit Setup added by Sasq
+@--------------------------------------------------------------------------------
+	@ Disable cache
+	mov	r0, #0
+	mcr	p15, 0, r0, c7, c5, 0		@ Instruction cache
+	mcr	p15, 0, r0, c7, c6, 0		@ Data cache
+
+	@ Wait for write buffer to empty 
+	mcr	p15, 0, r0, c7, c10, 4
+
+	ldr	r0, =__dtcm_start
+	orr	r0,r0,#0x0a
+	mcr	p15, 0, r0, c9, c1,0		@ DTCM base = __dtcm_start, size = 16 KB
+
+	mov r0,#0x20
+	mcr	p15, 0, r0, c9, c1,1		@ ITCM base = 0 , size = 32 MB
+
+@--------------------------------------------------------------------------------
+@ Setup memory regions similar to Release Version
+@--------------------------------------------------------------------------------
+
+	@------------------------------------------------------------------------
+	@ Region 0 - IO registers
+	@------------------------------------------------------------------------
+	ldr	r0,=( PAGE_64M | 0x04000000 | 1)	
+	mcr	p15, 0, r0, c6, c0, 0
+
+	@------------------------------------------------------------------------
+	@ Region 1 - Main Memory
+	@------------------------------------------------------------------------
+	ldr	r0,=( PAGE_4M | 0x02000000 | 1)	
+	mcr	p15, 0, r0, c6, c1, 0
+
+	@------------------------------------------------------------------------
+	@ Region 2 - alternate vector base
+	@------------------------------------------------------------------------
+	ldr	r0,=( PAGE_4K | 0x00000000 | 1)	
+	mcr	p15, 0, r0, c6, c2, 0
+
+	@------------------------------------------------------------------------
+	@ Region 3 - DS Accessory (GBA Cart)
+	@------------------------------------------------------------------------
+	ldr	r0,=( PAGE_128M | 0x08000000 | 1)	
+	mcr	p15, 0, r0, c6, c3, 0
+
+	@------------------------------------------------------------------------
+	@ Region 4 - DTCM
+	@------------------------------------------------------------------------
+	ldr	r0,=__dtcm_start
+	orr	r0,r0,#(PAGE_16K | 1)
+	mcr	p15, 0, r0, c6, c4, 0
+
+	@------------------------------------------------------------------------
+	@ Region 5 - ITCM
+	@------------------------------------------------------------------------
+	ldr	r0,=__itcm_start
+
+	@ align to 32k boundary
+	mov	r0,r0,lsr #15
+	mov	r0,r0,lsl #15
+
+	orr	r0,r0,#(PAGE_32K | 1)
+	mcr	p15, 0, r0, c6, c5, 0
+
+	@------------------------------------------------------------------------
+	@ Region 6 - System ROM
+	@------------------------------------------------------------------------
+	ldr	r0,=( PAGE_32K | 0xFFFF0000 | 1)	
+	mcr	p15, 0, r0, c6, c6, 0
+
+	@------------------------------------------------------------------------
+	@ Region 7 - non cacheable main ram
+	@------------------------------------------------------------------------
+	ldr	r0,=( PAGE_4M  | 0x02400000 | 1)	
+	mcr	p15, 0, r0, c6, c7, 0
+
+	@------------------------------------------------------------------------
+	@ Write buffer enable
+	@------------------------------------------------------------------------
+	ldr	r0,=0b00000010
+	mcr	p15, 0, r0, c3, c0, 0
+#endif
+
+	@------------------------------------------------------------------------
+	@ DCache & ICache enable
+	@------------------------------------------------------------------------
+	ldr	r0,=0b01000010
+#if SAFE
+	mcr	p15, 0, r0, c2, c0, 0
+#endif
+	mcr	p15, 0, r0, c2, c0, 1
+
+#if SAFE
+	@------------------------------------------------------------------------
+	@ IAccess
+	@------------------------------------------------------------------------
+	ldr	r0,=0x36636633
+	mcr	p15, 0, r0, c5, c0, 3
+
+	@------------------------------------------------------------------------
+	@ DAccess
+	@------------------------------------------------------------------------
+	ldr	r0,=0x36333633
+	mcr     p15, 0, r0, c5, c0, 2
+#endif
+
+	@------------------------------------------------------------------------
+	@ Enable ICache, DCache, ITCM & DTCM
+	@------------------------------------------------------------------------
+	mrc	p15, 0, r0, c1, c0, 0
+#if SAFE
+	ldr	r1,= ITCM_ENABLE | DTCM_ENABLE | ICACHE_ENABLE | DCACHE_ENABLE | PROTECT_ENABLE
+#else
+	ldr	r1, =ICACHE_ENABLE|DCACHE_ENABLE|PROTECT_ENABLE
+#endif
+	orr	r0,r0,r1
+	mcr	p15, 0, r0, c1, c0, 0
+
+#if SAFE
+	mov	r0, #0x12		@ Switch to IRQ Mode
+	msr	cpsr, r0
+	ldr	sp, =__sp_irq		@ Set IRQ stack
+
+	mov	r0, #0x13		@ Switch to SVC Mode
+	msr	cpsr, r0
+	ldr	sp, =__sp_svc		@ Set SVC stack
+
+	mov	r0, #0x1F		@ Switch to System Mode
+	msr	cpsr, r0
+	ldr	sp, =__sp_usr		@ Set user stack
+#endif
+
+	ldr	r3, =Main
+	blx	r3			@ jump to user code
--- a/demo/src9/src/init.s
+++ b/demo/src9/src/init.s
@@ -0,0 +1,52 @@
+	.arch armv5te
+	.thumb
+
+	.global Init
+	.text
+	.thumb_func
+	.type Init, %function
+Init:
+	@; POWERCNT = POWER_ALL
+	ldr r0, =0x04000304
+	ldr r1, =0x20F
+	strh r1, [r0]
+
+	@; GFX_CLEAR_COLOR = black, opaque, id=63
+	ldr r0, =0x04000350
+	ldr r1, =0x3F1F0000
+	str r1, [r0]
+
+	@; GFX_CLEAR_DEPTH = 0x7FFF
+	ldr r1, =0x7FFF
+	strh r1, [r0, #4]
+
+	@; DISPCNT = MODE 0, BG0, 3D
+	ldr r0, =0x04000000
+	ldr r1, =0x10000|(1<<8)|(1<<3)
+	str r1, [r0]
+
+	@; GFX_CONTROL = ANTIALIAS, BLEND
+	ldr r0, =0x04000060
+	ldr r1, =(1<<3)|(1<<4)
+	strh r1, [r0]
+
+	@; DISPCNT_SUB = MODE 5, BG3, SPR, SPR_2D_BMP_256
+	ldr r0, =0x04001000
+	ldr r1, =0x10005|(1<<11)|(1<<12)|(2<<4)
+	str r1, [r0]
+
+	@; BG3CNT_SUB = BMP16 256x256
+	ldr r1, =0x4084
+	strh r1, [r0, #0x0E]
+
+	@; BG3PA_SUB = 256
+	ldr r1, =256
+	strh r1, [r0, #0x30]
+
+	@; BG3PD_SUB = 256
+	strh r1, [r0, #0x36]
+
+	bx lr
+
+	.align 2
+	.pool
--- a/demo/src9/src/main.c
+++ b/demo/src9/src/main.c
@@ -0,0 +1,573 @@
+#include <nds.h>
+
+#include "options.h"
+#include "trig.h"
+#include "3D.h"
+
+#define FADES 1 // ~128 bytes... (packed)
+
+#if FADES
+#define BRIGHT (1<<14)
+#define DARK (2<<14)
+
+int fade_current = 31 << 2; // starts white
+int fade_target = 16 << 2;
+#endif
+
+u32 cubelist[] =
+{
+	COMMAND(CDIFAMB, CBEGIN, CNORMAL, CVERTEX),
+	DIFAMB(RGB15(16, 16, 16), RGB15(8, 8, 8)),
+	QUAD,
+	NORMAL(0, 512, 0),
+	VERTEX(-64,-64,-64),
+
+	COMMAND(CVERTEX, CVERTEX, CVERTEX, CNORMAL),
+	VERTEX( 64,-64,-64),
+	VERTEX( 64,-64, 64),
+	VERTEX(-64,-64, 64),
+	NORMAL(0, 511, 0),
+
+	COMMAND(CVERTEX, CVERTEX, CVERTEX, CVERTEX),
+	VERTEX(-64, 64,-64),
+	VERTEX(-64, 64, 64),
+	VERTEX( 64, 64, 64),
+	VERTEX( 64, 64,-64),
+
+	COMMAND(CNORMAL, CVERTEX, CVERTEX, CVERTEX),
+	NORMAL(512, 0, 0),
+	VERTEX(-64,-64,-64),
+	VERTEX(-64,-64, 64),
+	VERTEX(-64, 64, 64),
+
+	COMMAND(CVERTEX, CNORMAL, CVERTEX, CVERTEX),
+	VERTEX(-64, 64,-64),
+	NORMAL(511, 0, 0),
+	VERTEX( 64,-64,-64),
+	VERTEX( 64, 64,-64),
+
+	COMMAND(CVERTEX, CVERTEX, CNORMAL, CVERTEX),
+	VERTEX( 64, 64, 64),
+	VERTEX( 64,-64, 64),
+	NORMAL( 0, 0, 512),
+	VERTEX(-64,-64,-64),
+
+	COMMAND(CVERTEX, CVERTEX, CVERTEX, CNORMAL),
+	VERTEX(-64, 64,-64),
+	VERTEX( 64, 64,-64),
+	VERTEX( 64,-64,-64),
+	NORMAL( 0, 0, 511),
+
+	COMMAND(CVERTEX, CVERTEX, CVERTEX, CVERTEX),
+	VERTEX(-64,-64, 64),
+	VERTEX( 64,-64, 64),
+	VERTEX( 64, 64, 64),
+	VERTEX(-64, 64, 64),
+};
+#define cube() call_list(cubelist, sizeof(cubelist)/sizeof(u32))
+
+u32 initlist[] =
+{
+	COMMAND(CVIEWPORT, CMTXMODE, CMTXLOAD4x4, 0),
+	VIEWPORT(0, 0, 255, 191), // full screen
+	PROJECTION, // matrix mode
+	// Projection matrix Perspective(fov=50<35>, near=0.1, far=40, ratio=1.33)
+	6587, 	0, 	0,	0,
+	0, 	8783, 	0,	0,
+	0, 	0, 	-4116,	-4096,
+	0, 	0, 	-821,	0,
+
+	COMMAND(CMTXMODE, CMTXIDENTITY, CLIGHTVECTOR, CLIGHTCOLOR),
+	MODELVIEW, // matrix mode
+	NORMAL(-96, -144, -482), // light vector
+	RGB15(31, 31, 31), // light color
+};
+#define init() call_list(initlist, sizeof(initlist)/sizeof(u32))
+
+void rotate(int angle, int v)
+{
+	int s = sin(angle);
+	int c = cos(angle);
+
+	if ( v == 1 ) {
+		MATRIX_MULT3x3 = 1<<12;
+		MATRIX_MULT3x3 = 0;
+		MATRIX_MULT3x3 = 0;
+
+		MATRIX_MULT3x3 = 0;
+		MATRIX_MULT3x3 = c;
+		MATRIX_MULT3x3 = s;
+
+		MATRIX_MULT3x3 = 0;
+		MATRIX_MULT3x3 = -s;
+		MATRIX_MULT3x3 = c;
+	} else if ( v == 2 ) {
+		MATRIX_MULT3x3 = c;
+		MATRIX_MULT3x3 = 0;
+		MATRIX_MULT3x3 = -s;
+
+		MATRIX_MULT3x3 = 0;
+		MATRIX_MULT3x3 = 1<<12;
+		MATRIX_MULT3x3 = 0;
+
+		MATRIX_MULT3x3 = s;
+		MATRIX_MULT3x3 = 0;
+		MATRIX_MULT3x3 = c;
+	} else {
+		MATRIX_MULT3x3 = c;
+		MATRIX_MULT3x3 = s;
+		MATRIX_MULT3x3 = 0;
+
+		MATRIX_MULT3x3 = -s;
+		MATRIX_MULT3x3 = c;
+		MATRIX_MULT3x3 = 0;
+
+		MATRIX_MULT3x3 = 0;
+		MATRIX_MULT3x3 = 0;
+		MATRIX_MULT3x3 = 1<<12;
+	}
+}
+
+void arms(int t, int m)
+{
+	push();
+//		translate(0.24f*m, 0.4f, 0);
+		translate(983*m, 1638, 0);
+//		rotateZ(30<33>*m);
+		rotateZ(2145*m);
+//		rotateX(cos(t)*20<32>+180<38>);
+		rotateX(((cos(t)*1430)>>12)+12868);
+//		translate(0, 0.3, 0);
+		translate(0, 1229, 0);
+		push();
+//			scale(0.1, 0.3, 0.1);
+			scale(410, 1229, 410);
+			cube();
+		pop();
+
+//		translate(0, 0.22, 0);
+		translate(0, 901, 0);
+//		rotateY(abs(cos(t))*16*m);
+		int act = cos(t)*1144;
+		if ( act < 0 )
+			act = -act;
+		rotateY((act>>12)*m);
+//		rotateX(100<30>);
+		rotateX(7149);
+//		rotateZ(165<36>*m);
+		rotateZ(11796*m);
+//		translate(0, 0.21, 0);
+		translate(0, 860, 0);
+		push();
+//			scale(0.09, 0.27, 0.09);
+			scale(369, 1106, 369);
+			cube();
+		pop();
+	pop();
+}
+
+void legs(int t, int m)
+{
+	push();
+//		translate(0.19*m, 0, -0.015);
+		translate(778*m, 0, -61);
+//		rotateX(cos(t)*13<31>-2.5);
+		rotateX(((cos(t)*929)>>12)-228); // 929->1024 = -12 bytes !
+
+//		translate(0, -0.7, 0);
+		translate(0, -2867, 0);
+		push();
+//			scale(0.1, 0.25, 0.1);
+			scale(410, 1024, 410);
+			cube();
+		pop();
+
+//		float d = -min(0, cos(t+3.14157/2));
+		int d = -cos(t+6434);
+		if ( d < 0 )
+			d = 0;
+//		translate(0, d*0.15-0.52, d*0.015-0.06);
+		translate(0, ((d*512)>>12)-2130, ((d*61)>>12)-246);
+//		rotateX(10<31>);
+		rotateX(715);
+		push();
+//			scale(0.1, 0.3, 0.1);
+			scale(410, 1229, 410);
+			cube();
+		pop();
+
+//		translate(0, -0.3, 0.1);
+		translate(0, -1229, 410);
+		push();
+//			scale(0.1, 0.03, 0.2);
+			scale(410, 123, 819);
+			cube();
+		pop();
+	pop();
+}
+
+void robot(int t, int T)
+{
+#define robot_fall_speed 2048
+#define robot_start 40
+	t *= 402;
+
+	// ast = |sin(t)|
+	int ast = sin(t);
+	if ( ast < 0 )
+		ast = -ast;
+
+	// act = |cos(t)|
+	int act = cos(t);
+	if ( act < 0 )
+		act = -act;
+
+	int fall;
+
+	// head
+	push();
+//		translate(0, 0.71+0.05*abs(sin(t)), 0.04);
+		fall = (robot_start+88)*robot_fall_speed - T*robot_fall_speed;
+		if ( fall < 0 )
+			fall = 0;
+		translate(0, 2908+((ast*204)>>12)+fall, 164);
+//		rotateX(-5<>+10<31>*abs(cos(t)));
+		rotateX((-357+715*act)>>12);
+//		scale(0.2, 0.2, 0.2);
+		scale(819, 819, 819);
+		cube();
+	pop();
+
+	// body
+	push();
+//		translate(0, 0.05*abs(sin(t)), 0);
+		fall = (robot_start+48)*robot_fall_speed - T*robot_fall_speed;
+		if ( fall < 0 )
+			fall = 0;
+		translate(0, ((ast*204)>>12)+fall, 0);
+//		rotateY(2.5<EFBFBD>-5<>*cos(t));
+		rotateY(179-((357*cos(t))>>12));
+//		scale(0.3, 0.5, 0.17);
+		scale(1229, 2048, 696);
+		cube();
+	pop();
+
+	// left side
+	push();
+	fall = (robot_start+56)*robot_fall_speed - T*robot_fall_speed;
+	if ( fall < 0 )
+		fall = 0;
+	translate(0, fall, 0);
+	arms(t, -1);
+	pop();
+
+	push();
+	fall = (robot_start+8)*robot_fall_speed - T*robot_fall_speed;
+	if ( fall < 0 )
+		fall = 0;
+	translate(0, fall, 0);
+	legs(t, -1);
+	pop();
+
+	// right side
+	t += 12868;
+	push();
+	fall = (robot_start+64)*robot_fall_speed - T*robot_fall_speed;
+	if ( fall < 0 )
+		fall = 0;
+	translate(0, fall, 0);
+	arms(t, 1);
+	pop();
+
+	push();
+	fall = (robot_start+24)*robot_fall_speed - T*robot_fall_speed;
+	if ( fall < 0 )
+		fall = 0;
+	translate(0, fall, 0);
+	legs(t, 1);
+	pop();
+}
+
+// r * cos((x*pi/2)/r) * cos((y*pi/2)/r) - r
+// returns a fixed point altitude value in fx.6
+int f(int x, int y)
+{
+	const int r = 32;
+//	const int _r = (int)((0.5f/(float)r)*64.0f); // == 1 ...
+//	x = (x*_r);
+//	y = (y*_r);
+	return ((r * (uber_cos(x) * uber_cos(y))>>18)) - (r<<6);
+}
+
+void sonic(int t)
+{
+	int i = 0;
+	int y, x;
+
+	rotateX(-1024);
+
+	if ( t < 256 )
+		t = 0;
+	else
+		t = -(t*2)&127;
+
+	for ( y = -7*64+t ; y < 6*64+t ; y += 64 ) {
+		for ( x = -7*64 ; x < 6*64 ; x += 64 ) {
+			GFX_VERTEX10 = VERTEX(x,    f(x,    y),    y);
+			GFX_VERTEX10 = VERTEX(x,    f(x,    y+64), y+64);
+			GFX_VERTEX10 = VERTEX(x+64, f(x+64, y+64), y+64);
+			GFX_VERTEX10 = VERTEX(x+64, f(x+64, y),    y);
+			GFX_COLOR = (i&1) ? RGB15(29, 15, 4) : RGB15(4, 11, 28);
+			i++;
+		}
+		//i++;
+	}
+}
+
+void spot(int c1, int c2, int w, int h)
+{
+	GFX_BEGIN = QUAD;
+	GFX_COLOR = c1;
+	GFX_VERTEX10 = VERTEX(-64, h, 0);
+	GFX_VERTEX10 = VERTEX( 64, h, 0);
+	GFX_COLOR = c2;
+	GFX_VERTEX10 = VERTEX( 64+w, 64, 0);
+	GFX_VERTEX10 = VERTEX(-64-w, 64, 0);
+}
+
+void robot_solo(int t)
+{
+	int T = t;
+	if ( t < 256 ) {
+		// stopped for the intro
+		t = 0;
+	} else if ( t > 1152 ) {
+		// tracted by the flying saucer
+		int y;
+		y = (t-1152)*58;
+		if ( y > 22528 )
+			y = 22528;
+		translate(0, y, 0);
+		y = (8*4096-y)/8;
+		scale(y, y, y);
+		t = 0;
+	}
+	robot(t, T);
+}
+
+int jump(int t)
+{
+	if ( t < 512 )
+		return 0;
+	t = t%64;
+	t = t*8-384+128;
+	t = 3072-t*t;
+	if ( t < 0 )
+		return 0;
+	return t;
+}
+
+void robot_clone_wars(int t)
+{
+#define SPACE_FLOOR 1
+#define FINAL_WALK_ON 1
+#define LINES 6
+#define SPACING 16384
+	int i, j;
+
+	//rotateX(2560);
+	if ( (t & 1) == 0 ) { // different camera for different screens
+		translate(5120, 44032, 20480);
+		rotateZ(1792);
+	} else {
+		rotateX(2560);
+	}
+
+#if SPACE_FLOOR
+	push();
+	translate(0, -6144, -4096*4);
+	rotateX(-12867/2);
+	scale(4096*16, 4096*10, 4096*10);
+	spot(RGB15(0, 7, 3), RGB15(0, 2, 1), 0, -64);
+	pop();
+#endif
+
+#if SPACE_FLOOR
+	int z = t*256-16384;
+#else
+	int z = t*256*8-16384*8;
+#endif
+	if ( z > 0 )
+		z = 0;
+
+	int lines = (t-192+8)/8;
+
+	if ( lines < 2 )
+		lines = 2;
+	else if ( lines > LINES )
+		lines = LINES;
+
+	for ( i = 1 ; i < lines ; i++ ) {
+		translate(0, 0, -8192);
+		for ( j = 0 ; j < i ; j++ ) {
+			if ( j == 2 && i == 5 )
+				continue;
+			push();
+				translate(j * SPACING - ((i-1)*SPACING)/2, jump(t-i*4)+z, 0);
+#if FINAL_WALK_ON
+				if ( t < 512 )
+					robot(8, 256);
+				else
+					robot(8+t-512, 256);
+#else
+				robot(8);
+#endif
+			pop();
+		}
+	}
+}
+
+void background(int t)
+{
+#if FADES
+	if ( t == 1536 ) {
+		fade_current = 0;
+		fade_target = 16<<2;
+	}
+#endif
+	if ( t < 1536 )
+		spot(RGB15(31, 31, 31), RGB15(12, 12, 31), 0, -64);
+	else
+		spot(RGB15(1, 5, 2), RGB15(0, 0, 0), 0, -64);
+}
+
+void fs(int t)
+{
+	static int h[9] = { 64,		58,		51,		32,		26,		19,		10,		1,		0 };
+	static int r[9] = { 0,		819,	1229,	1638,	3276,	4096,	3276,	2048,	0 };
+
+	int y = 1024*32-t*32;
+	if ( y < 0 )
+		y = 0;
+	translate(y, y, -y);
+
+	GFX_COLOR = RGB15(15, 15, 15);
+	int i, j;
+	for ( i = 0 ; i < 8 ; i++ ) {
+		for ( j = 0 ; j < 8192 ; j += 512 ) {
+			GFX_NORMAL = VERTEX((uber_cos(j) * r[i])>>18, h[i], (uber_sin(j) * r[i])>>18);
+			GFX_VERTEX10 = VERTEX((uber_cos(j) * r[i])>>18, h[i], (uber_sin(j) * r[i])>>18);
+			GFX_VERTEX10 = VERTEX((uber_cos(j+512) * r[i])>>18, h[i], (uber_sin(j+512) * r[i])>>18);
+			GFX_VERTEX10 = VERTEX((uber_cos(j+512) * r[i+1])>>18, h[i+1], (uber_sin(j+512) * r[i+1])>>18);
+			GFX_VERTEX10 = VERTEX((uber_cos(j) * r[i+1])>>18, h[i+1], (uber_sin(j) * r[i+1])>>18);
+		}
+	}
+}
+
+void fs_beam(int t)
+{
+	int h = 64+8+384-t*4;
+	if ( h < 0 )
+		h = 0;
+	GFX_POLY_FORMAT = LIGHT0|POLYFRONT|ALPHA(15);
+	spot(RGB15(0, 31, 4), RGB15(0, 31, 4), -32, h-384);
+}
+
+typedef struct {
+	void (*draw)(int time);
+	s16 start, end;
+	s8 ty, tz;
+	s8 s;
+} Drawable;
+
+Drawable demo[] =
+{
+	// draw					start, 	end		ty, tz		scale
+	{ robot_solo, 			0, 		1536,	 1, -15, 	4 	},
+	{ sonic, 				0, 		1536, 	-8, -24, 	4 	},
+	{ fs, 					0, 		1536, 	21, -15, 	4 	},
+	{ robot_clone_wars, 	1536, 	2576,	-4, -16, 	2 	},
+	{ background, 			0, 		2576, 	 9, -44, 	32 	},
+	{ fs_beam, 				1024, 	1536, 	18, -13, 	4 	},
+};
+
+void Main()
+{
+	extern void Init();
+	Init();
+	init();
+
+	// Init sprites to display like a big bitmap background
+	//VRAM_D_CR = VRAM_D_SUB_SPRITE | VRAM_ENABLE;
+	int y, x;
+	u16 * ptr = OAM_SUB;
+	for ( y = 0 ; y < 3 ; y++ ) {
+		for ( x = 0 ; x < 4 ; x++ ) {
+			// attribute 0
+			*ptr++ = ATTR0_BMP | (y*64);
+			// attribute 1
+			*ptr++ = ATTR1_SIZE_64 | (x*64);
+			// attribute 2
+			*ptr++ = ATTR2_ALPHA(15) | (x*8+y*256);
+			// pad (matrix)
+			ptr++;
+		}
+	}
+	
+	int t = 0;
+
+	while ( t < 2576 ) {
+#if FADES
+		if ( fade_current < fade_target )
+			fade_current++;
+		else if ( fade_current > fade_target )
+			fade_current--;
+
+		int v = 0;
+		if ( fade_current < (16 << 2) )
+			v = DARK | (16 - (fade_current >> 2));
+		else if ( fade_current > (16 << 2) )
+			v = BRIGHT | ((fade_current >> 2) - 16);
+
+		REG_MASTER_BRIGHT = v;
+		REG_MASTER_BRIGHT_SUB = v;
+#endif
+
+		identity();
+
+		// Dual screen 3D
+		if ( t&1 ) {
+			REG_POWERCNT &= ~POWER_SWAP_LCDS; // main on bottom
+			VRAM_C_CR = VRAM_ENABLE;
+			VRAM_D_CR = VRAM_D_SUB_SPRITE | VRAM_ENABLE;
+			REG_DISPCAPCNT = DCAP_BANK(2) | DCAP_ENABLE | DCAP_SIZE(3);
+		} else {
+			REG_POWERCNT |= POWER_SWAP_LCDS; // main on top
+			VRAM_C_CR = VRAM_C_SUB_BG | VRAM_ENABLE;
+			VRAM_D_CR = VRAM_ENABLE;
+			REG_DISPCAPCNT = DCAP_BANK(3) | DCAP_ENABLE | DCAP_SIZE(3);
+			translate(0, -5*4096, 0);
+		}
+
+		// Reset polygon attributes
+		GFX_POLY_FORMAT = LIGHT0|POLYFRONT|SOLID;
+
+		// Lightweight 3D player -olol
+		int i;
+		for ( i = 0 ; i < sizeof(demo)/sizeof(Drawable) ; i++ ) {
+			if ( t >= demo[i].start && t < demo[i].end ) {
+				push();
+					translate(0, demo[i].ty*1024, demo[i].tz*1024);
+					scale(demo[i].s*1024, demo[i].s*1024, demo[i].s*1024);
+					demo[i].draw(t-demo[i].start);
+				pop();
+			}
+		}
+
+		t++;
+		swap();
+	}
+
+	extern void seeya();
+	seeya();
+}
--- a/demo/src9/src/msg
+++ b/demo/src9/src/msg
--- a/demo/src9/src/msg.s
+++ b/demo/src9/src/msg.s
@@ -0,0 +1,68 @@
+	.arch armv5te
+	.align 2
+	.text
+	.thumb
+	.global seeya
+	.thumb_func
+	.type seeya, %function
+seeya:
+	@; dispcnt = mode 0, bg0 active
+	ldr r0, =0x04000000
+	ldr r1, =0x10000 | (1<<8)
+	str r1, [r0]
+
+	@; bg0cnt = tilebase=1, mapbase=0
+	ldr r1, =(1<<2)
+	str r1, [r0, #8]
+
+	@; scroll text horizontally
+	ldr r1, =256-108
+	strh r1, [r0, #0x10]
+
+	@; scroll text vertically
+	ldr r1, =256-92
+	strh r1, [r0, #0x12]
+
+	@; enable vram A in bg mode @0x06000000
+	ldr r0, =0x04000240
+	ldr r1, =(1<<7)|1
+	str r1, [r0]
+
+	@; bg palette color 1 = white
+	ldr r0, =0x05000002
+	ldr r1, =(1<<15)-1
+	strh r1, [r0]
+
+	@; fill map
+	ldr r0, =0x06000000
+	mov r1, #0
+textloop:
+	strh r1, [r0]
+	add r0, #2
+	add r1, #1
+	cmp r1, #(msgend-msg)/8
+	bne textloop
+
+	@; unpack tiles directly to vram
+	ldr r0, =msg
+	ldr r1, =0x06004000
+	ldr r2, =unpackinfo
+	swi 0x10
+
+	@; infinite wait (interrupts aren't enabled ;)
+	swi 0x05
+
+	.align 2
+	.pool
+
+	.align 2
+msg:
+	.incbin "msg"
+msgend:
+
+	.align 2
+unpackinfo:
+	.hword msgend-msg
+	.byte 1
+	.byte 4
+	.word 0
--- a/demo/src9/src/options.h
+++ b/demo/src9/src/options.h
@@ -0,0 +1,8 @@
+#ifndef _OPTIONS_H_
+#define _OPTIONS_H_
+
+#define STDLIB 0
+#define CAMERA 0
+#define SAFE 0
+
+#endif // _OPTIONS_H_
--- a/demo/src9/src/trig.c
+++ b/demo/src9/src/trig.c
@@ -0,0 +1,12 @@
+#include "trig.h"
+#include <nds.h>
+
+int uber_sin(int x)
+{
+	// TODO this may be simplified
+	x = 4096 - (x & 0x1FFF);
+	int ax = x;
+	if ( ax < 0 )
+		ax = -ax;
+	return 4 * x - ((4 * x * ax) >> 12);
+}
--- a/demo/src9/src/trig.h
+++ b/demo/src9/src/trig.h
@@ -0,0 +1,17 @@
+#ifndef _TRIG_H_
+#define _TRIG_H_
+
+// approximation of sin with pi=4096
+//
+// pi-(x mod 2pi) then
+// 4.x - 4.x.|x|
+int uber_sin(int x);
+#define uber_cos(x) uber_sin(x + 0x800)
+
+// Conversion from old cos/sin to new uber cos/sin
+// The old was working with real value of pi in fixed point 12 (pi = 12867)
+// while the new one uses pi = 4096
+#define sin(x) uber_sin((((x) * 325) >> 10))
+#define cos(x) uber_cos((((x) * 325) >> 10))
+
+#endif // _TRIG_H_