DevKit7/libctru
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

rewrote GPU examples/gpu/data/test.vsh

Browse Source
This commit is contained in:
smea 2014-11-19 17:13:43 -08:00
parent 5ecb358b56
commit c1ad708045
8 changed files with 1016 additions and 200 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

52
examples/gpu/data/test.vsh
View File

@ -1,7 +1,7 @@
; make sure you update aemstro_as for this (27/05/14) ; make sure you update aemstro_as for this (15/11/14)
; setup constants ; setup constants
.const 5, 0.0, 0.0, -0.99, 1.0 .const 20, 1.0, 0.0, 0.5, 1.0
; setup outmap ; setup outmap
.out o0, result.position .out o0, result.position
@ -10,34 +10,36 @@
.out o3, result.texcoord1 .out o3, result.texcoord1
.out o4, result.texcoord2 .out o4, result.texcoord2
; setup uniform map (not required) ; setup uniform map (required to use SHDR_GetUniformRegister)
.uniform 0x10, 0x13, mdlvMtx .uniform 0, 3, projection ; c0-c3 = projection matrix
.uniform 0x14, 0x17, projMtx .uniform 4, 7, modelview ; c4-c7 = modelview matrix
.uniform 8, 8, lightDirection ; c8 = light direction vector
.uniform 9, 9, lightAmbient ; c9 = light ambient color
;code ;code
main: main:
mov d1A, d00 (0x4) mov r1, v0 (0x6)
mov d1A, d25 (0x3) mov r1, c20 (0x3)
; tempreg = mdlvMtx * in.pos ; tempreg = mdlvMtx * in.pos
dp4 d10, d44, d1A (0x0) dp4 r0, c4, r1 (0x0)
dp4 d10, d45, d1A (0x1) dp4 r0, c5, r1 (0x1)
dp4 d10, d46, d1A (0x2) dp4 r0, c6, r1 (0x2)
mov d10, d25 (0x3) mov r0, c20 (0x3)
; result.pos = projMtx * tempreg ; result.pos = projMtx * tempreg
dp4 d00, d40, d10 (0x0) dp4 o0, c0, r0 (0x0)
dp4 d00, d41, d10 (0x1) dp4 o0, c1, r0 (0x1)
dp4 d00, d42, d10 (0x2) dp4 o0, c2, r0 (0x2)
dp4 d00, d43, d10 (0x3) dp4 o0, c3, r0 (0x3)
; result.texcoord = in.texcoord ; result.texcoord = in.texcoord
mov d02, d01 (0x5) mov o2, v1 (0x5)
mov d03, d25 (0x7) mov o3, c20 (0x7)
mov d04, d25 (0x7) mov o4, c20 (0x7)
; result.color = crappy lighting ; result.color = crappy lighting
dp3 d1A, d44, d02 (0x0) dp3 r0, c8, v2 (0x6)
dp3 d1A, d45, d02 (0x1) max r0, c20, r0 (0x4)
dp3 d1A, d46, d02 (0x2) mul r0, c9, r0 (0x8)
dp4 d01, d00, d1A (0x6) add o1, c9, r0 (0x6)
mov d01, d25 (0x3) mov o1, c20 (0x3)
flush flush
end end
endmain: endmain:
@ -47,8 +49,8 @@
.opdesc _y__, xyzw, xyzw ; 0x1 .opdesc _y__, xyzw, xyzw ; 0x1
.opdesc __z_, xyzw, xyzw ; 0x2 .opdesc __z_, xyzw, xyzw ; 0x2
.opdesc ___w, xyzw, xyzw ; 0x3 .opdesc ___w, xyzw, xyzw ; 0x3
.opdesc xyz_, xyzw, xyzw ; 0x4 .opdesc xyz_, yyyy, xyzw ; 0x4
.opdesc xyzw, xyzw, xyzw ; 0x5 .opdesc xyzw, xyzw, xyzw ; 0x5
.opdesc xyz_, xyzw, xyzw ; 0x6 .opdesc xyz_, xyzw, xyzw ; 0x6
.opdesc xyzw, yyyw, xyzw ; 0x7 .opdesc xyzw, yyyw, xyzw ; 0x7
.opdesc xyzw, wwww, wwww ; 0x8 .opdesc xyz_, wwww, xyzw ; 0x8

BIN
examples/gpu/data/texture.bin Normal file
View File

Binary file not shown.

16
examples/gpu/source/_gs.s Normal file
View File

@ -0,0 +1,16 @@
.section ".text"
.arm
.align 4
.global _vboMemcpy50
# r0 : dst
# r1 : src
# fixed size 0x50
_vboMemcpy50:
push {r4-r11}
ldmia r1!, {r2-r12}
stmia r0!, {r2-r12}
ldmia r1!, {r2-r12}
stmia r0!, {r2-r12}
pop {r4-r11}
bx lr

431
examples/gpu/source/gs.c Normal file
View File

@ -0,0 +1,431 @@
#include <stdlib.h>
#include <string.h>
#include <malloc.h>
#include <3ds.h>
#include "gs.h"
#include "math.h"
#define BUFFERMATRIXLIST_SIZE (GS_MATRIXSTACK_SIZE*4)
static void gsInitMatrixStack();
Handle linearAllocMutex;
static u32 gsMatrixStackRegisters[GS_MATRIXTYPES];
typedef struct
{
u32 offset;
mtx44 data;
}bufferMatrix_s;
bufferMatrix_s bufferMatrixList[BUFFERMATRIXLIST_SIZE];
int bufferMatrixListLength;
//----------------------
// GS SYSTEM STUFF
//----------------------
void initBufferMatrixList()
{
bufferMatrixListLength=0;
}
void gsInit(DVLB_s* shader)
{
gsInitMatrixStack();
initBufferMatrixList();
svcCreateMutex(&linearAllocMutex, false);
if(shader)
{
gsMatrixStackRegisters[0]=SHDR_GetUniformRegister(shader, "projection", 0);
gsMatrixStackRegisters[1]=SHDR_GetUniformRegister(shader, "modelview", 0);
}
}
void gsExit(void)
{
svcCloseHandle(linearAllocMutex);
}
void gsStartFrame(void)
{
GPUCMD_SetBufferOffset(0);
initBufferMatrixList();
}
void* gsLinearAlloc(size_t size)
{
void* ret=NULL;
svcWaitSynchronization(linearAllocMutex, U64_MAX);
ret=linearAlloc(size);
svcReleaseMutex(linearAllocMutex);
return ret;
}
void gsLinearFree(void* mem)
{
svcWaitSynchronization(linearAllocMutex, U64_MAX);
linearFree(mem);
svcReleaseMutex(linearAllocMutex);
}
//----------------------
// MATRIX STACK STUFF
//----------------------
static mtx44 gsMatrixStacks[GS_MATRIXTYPES][GS_MATRIXSTACK_SIZE];
static u32 gsMatrixStackRegisters[GS_MATRIXTYPES]={0x00, 0x04};
static u8 gsMatrixStackOffsets[GS_MATRIXTYPES];
static bool gsMatrixStackUpdated[GS_MATRIXTYPES];
static GS_MATRIX gsCurrentMatrixType;
static void gsInitMatrixStack()
{
int i;
for(i=0; i<GS_MATRIXTYPES; i++)
{
gsMatrixStackOffsets[i]=0;
gsMatrixStackUpdated[i]=true;
loadIdentity44((float*)gsMatrixStacks[i][0]);
}
gsCurrentMatrixType=GS_PROJECTION;
}
float* gsGetMatrix(GS_MATRIX m)
{
if(m<0 || m>=GS_MATRIXTYPES)return NULL;
return (float*)gsMatrixStacks[m][gsMatrixStackOffsets[m]];
}
int gsLoadMatrix(GS_MATRIX m, float* data)
{
if(m<0 || m>=GS_MATRIXTYPES || !data)return -1;
memcpy(gsGetMatrix(m), data, sizeof(mtx44));
gsMatrixStackUpdated[m]=true;
return 0;
}
int gsPushMatrix()
{
const GS_MATRIX m=gsCurrentMatrixType;
if(m<0 || m>=GS_MATRIXTYPES)return -1;
if(gsMatrixStackOffsets[m]<0 || gsMatrixStackOffsets[m]>=GS_MATRIXSTACK_SIZE-1)return -1;
float* cur=gsGetMatrix(m);
gsMatrixStackOffsets[m]++;
memcpy(gsGetMatrix(m), cur, sizeof(mtx44));
return 0;
}
int gsPopMatrix()
{
const GS_MATRIX m=gsCurrentMatrixType;
if(m<0 || m>=GS_MATRIXTYPES)return -1;
if(gsMatrixStackOffsets[m]<1 || gsMatrixStackOffsets[m]>=GS_MATRIXSTACK_SIZE)return -1;
gsMatrixStackOffsets[m]--;
gsMatrixStackUpdated[m]=true;
return 0;
}
int gsMatrixMode(GS_MATRIX m)
{
if(m<0 || m>=GS_MATRIXTYPES)return -1;
gsCurrentMatrixType=m;
return 0;
}
//------------------------
// MATRIX TRANSFORM STUFF
//------------------------
int gsMultMatrix(float* data)
{
if(!data)return -1;
mtx44 tmp;
multMatrix44(gsGetMatrix(gsCurrentMatrixType), data, (float*)tmp);
memcpy(gsGetMatrix(gsCurrentMatrixType), (float*)tmp, sizeof(mtx44));
gsMatrixStackUpdated[gsCurrentMatrixType]=true;
return 0;
}
void gsLoadIdentity()
{
loadIdentity44(gsGetMatrix(gsCurrentMatrixType));
gsMatrixStackUpdated[gsCurrentMatrixType]=true;
}
void gsProjectionMatrix(float fovy, float aspect, float near, float far)
{
initProjectionMatrix(gsGetMatrix(gsCurrentMatrixType), fovy, aspect, near, far);
gsMatrixStackUpdated[gsCurrentMatrixType]=true;
}
void gsRotateX(float x)
{
rotateMatrixX(gsGetMatrix(gsCurrentMatrixType), x, false);
gsMatrixStackUpdated[gsCurrentMatrixType]=true;
}
void gsRotateY(float y)
{
rotateMatrixY(gsGetMatrix(gsCurrentMatrixType), y, false);
gsMatrixStackUpdated[gsCurrentMatrixType]=true;
}
void gsRotateZ(float z)
{
rotateMatrixZ(gsGetMatrix(gsCurrentMatrixType), z, false);
gsMatrixStackUpdated[gsCurrentMatrixType]=true;
}
void gsScale(float x, float y, float z)
{
scaleMatrix(gsGetMatrix(gsCurrentMatrixType), x, y, z);
gsMatrixStackUpdated[gsCurrentMatrixType]=true;
}
void gsTranslate(float x, float y, float z)
{
translateMatrix(gsGetMatrix(gsCurrentMatrixType), x, y, z);
gsMatrixStackUpdated[gsCurrentMatrixType]=true;
}
//----------------------
// MATRIX RENDER STUFF
//----------------------
static void gsSetUniformMatrix(u32 startreg, float* m)
{
float param[16];
param[0x0]=m[3]; //w
param[0x1]=m[2]; //z
param[0x2]=m[1]; //y
param[0x3]=m[0]; //x
param[0x4]=m[7];
param[0x5]=m[6];
param[0x6]=m[5];
param[0x7]=m[4];
param[0x8]=m[11];
param[0x9]=m[10];
param[0xa]=m[9];
param[0xb]=m[8];
param[0xc]=m[15];
param[0xd]=m[14];
param[0xe]=m[13];
param[0xf]=m[12];
GPU_SetUniform(startreg, (u32*)param, 4);
}
static int gsUpdateTransformation()
{
GS_MATRIX m;
for(m=0; m<GS_MATRIXTYPES; m++)
{
if(gsMatrixStackUpdated[m])
{
if(m==GS_PROJECTION && bufferMatrixListLength<BUFFERMATRIXLIST_SIZE)
{
GPUCMD_GetBuffer(NULL, NULL, &bufferMatrixList[bufferMatrixListLength].offset);
memcpy(bufferMatrixList[bufferMatrixListLength].data, gsGetMatrix(m), sizeof(mtx44));
bufferMatrixListLength++;
}
gsSetUniformMatrix(gsMatrixStackRegisters[m], gsGetMatrix(m));
gsMatrixStackUpdated[m]=false;
}
}
return 0;
}
void gsAdjustBufferMatrices(mtx44 transformation)
{
int i;
u32* buffer;
u32 offset;
GPUCMD_GetBuffer(&buffer, NULL, &offset);
for(i=0; i<bufferMatrixListLength; i++)
{
u32 o=bufferMatrixList[i].offset;
if(o+2<offset) //TODO : better check, need to account for param size
{
mtx44 newMatrix;
GPUCMD_SetBufferOffset(o);
multMatrix44((float*)bufferMatrixList[i].data, (float*)transformation, (float*)newMatrix);
gsSetUniformMatrix(gsMatrixStackRegisters[GS_PROJECTION], (float*)newMatrix);
}
}
GPUCMD_SetBufferOffset(offset);
}
//----------------------
// VBO STUFF
//----------------------
int gsVboInit(gsVbo_s* vbo)
{
if(!vbo)return -1;
vbo->data=NULL;
vbo->currentSize=0;
vbo->maxSize=0;
vbo->commands=NULL;
vbo->commandsSize=0;
return 0;
}
int gsVboCreate(gsVbo_s* vbo, u32 size)
{
if(!vbo)return -1;
vbo->data=gsLinearAlloc(size);
vbo->numVertices=0;
vbo->currentSize=0;
vbo->maxSize=size;
return 0;
}
void* gsVboGetOffset(gsVbo_s* vbo)
{
if(!vbo)return NULL;
return (void*)(&((u8*)vbo->data)[vbo->currentSize]);
}
int gsVboAddData(gsVbo_s* vbo, void* data, u32 size, u32 units)
{
if(!vbo || !data || !size)return -1;
if(((s32)vbo->maxSize)-((s32)vbo->currentSize) < size)return -1;
memcpy(gsVboGetOffset(vbo), data, size);
vbo->currentSize+=size;
vbo->numVertices+=units;
return 0;
}
int gsVboFlushData(gsVbo_s* vbo)
{
if(!vbo)return -1;
//unnecessary if we use flushAndRun
// GSPGPU_FlushDataCache(NULL, vbo->data, vbo->currentSize);
return 0;
}
int gsVboDestroy(gsVbo_s* vbo)
{
if(!vbo)return -1;
if(vbo->commands)free(vbo->commands);
if(vbo->data)gsLinearFree(vbo->data);
gsVboInit(vbo);
return 0;
}
extern u32 debugValue[];
void GPU_DrawArrayDirectly(GPU_Primitive_t primitive, u8* data, u32 n)
{
//set attribute buffer address
GPUCMD_AddSingleParam(0x000F0200, (osConvertVirtToPhys((u32)data))>>3);
//set primitive type
GPUCMD_AddSingleParam(0x0002025E, primitive);
GPUCMD_AddSingleParam(0x0002025F, 0x00000001);
//index buffer not used for drawArrays but 0x000F0227 still required
GPUCMD_AddSingleParam(0x000F0227, 0x80000000);
//pass number of vertices
GPUCMD_AddSingleParam(0x000F0228, n);
GPUCMD_AddSingleParam(0x00010253, 0x00000001);
GPUCMD_AddSingleParam(0x00010245, 0x00000000);
GPUCMD_AddSingleParam(0x000F022E, 0x00000001);
GPUCMD_AddSingleParam(0x00010245, 0x00000001);
GPUCMD_AddSingleParam(0x000F0231, 0x00000001);
// GPUCMD_AddSingleParam(0x000F0111, 0x00000001); //breaks stuff
}
//not thread safe
int gsVboPrecomputeCommands(gsVbo_s* vbo)
{
if(!vbo || vbo->commands)return -1;
static u32 tmpBuffer[128];
u32* savedAdr; u32 savedSize, savedOffset;
GPUCMD_GetBuffer(&savedAdr, &savedSize, &savedOffset);
GPUCMD_SetBuffer(tmpBuffer, 128, 0);
GPU_DrawArrayDirectly(GPU_TRIANGLES, vbo->data, vbo->numVertices);
GPUCMD_GetBuffer(NULL, NULL, &vbo->commandsSize);
vbo->commands=memalign(0x4, vbo->commandsSize*4);
if(!vbo->commands)return -1;
memcpy(vbo->commands, tmpBuffer, vbo->commandsSize*4);
GPUCMD_SetBuffer(savedAdr, savedSize, savedOffset);
return 0;
}
extern u32* gpuCmdBuf;
extern u32 gpuCmdBufSize;
extern u32 gpuCmdBufOffset;
void _vboMemcpy50(u32* dst, u32* src);
void _GPUCMD_AddRawCommands(u32* cmd, u32 size)
{
if(!cmd || !size)return;
if(size*4==0x50)_vboMemcpy50(&gpuCmdBuf[gpuCmdBufOffset], cmd);
else memcpy(&gpuCmdBuf[gpuCmdBufOffset], cmd, size*4);
gpuCmdBufOffset+=size;
}
int gsVboDraw(gsVbo_s* vbo)
{
if(!vbo || !vbo->data || !vbo->currentSize || !vbo->maxSize)return -1;
gsUpdateTransformation();
gsVboPrecomputeCommands(vbo);
// u64 val=svcGetSystemTick();
if(vbo->commands)
{
_GPUCMD_AddRawCommands(vbo->commands, vbo->commandsSize);
}else{
GPU_DrawArrayDirectly(GPU_TRIANGLES, vbo->data, vbo->numVertices);
}
// debugValue[5]+=(u32)(svcGetSystemTick()-val);
// debugValue[6]++;
return 0;
}

59
examples/gpu/source/gs.h Normal file
View File

@ -0,0 +1,59 @@
#ifndef GS_H
#define GS_H
#include <3ds.h>
#include "math.h"
#define GS_MATRIXSTACK_SIZE (8)
typedef enum
{
GS_PROJECTION = 0,
GS_MODELVIEW = 1,
GS_MATRIXTYPES
}GS_MATRIX;
typedef struct
{
u8* data;
u32 currentSize; // in bytes
u32 maxSize; // in bytes
u32 numVertices;
u32* commands;
u32 commandsSize;
}gsVbo_s;
void gsInit(DVLB_s* shader);
void gsExit(void);
void gsStartFrame(void);
void gsAdjustBufferMatrices(mtx44 transformation);
void* gsLinearAlloc(size_t size);
void gsLinearFree(void* mem);
float* gsGetMatrix(GS_MATRIX m);
int gsLoadMatrix(GS_MATRIX m, float* data);
int gsPushMatrix();
int gsPopMatrix();
int gsMatrixMode(GS_MATRIX m);
void gsLoadIdentity();
void gsProjectionMatrix(float fovy, float aspect, float near, float far);
void gsRotateX(float x);
void gsRotateY(float y);
void gsRotateZ(float z);
void gsScale(float x, float y, float z);
void gsTranslate(float x, float y, float z);
int gsMultMatrix(float* data);
int gsVboInit(gsVbo_s* vbo);
int gsVboCreate(gsVbo_s* vbo, u32 size);
int gsVboFlushData(gsVbo_s* vbo);
int gsVboDestroy(gsVbo_s* vbo);
int gsVboDraw(gsVbo_s* vbo);
void* gsVboGetOffset(gsVbo_s* vbo);
int gsVboAddData(gsVbo_s* vbo, void* data, u32 size, u32 units);
#endif

404
examples/gpu/source/main.c
View File

@ -1,201 +1,325 @@
///////////////////////////////////////
// GPU example //
///////////////////////////////////////
//this example is meant to show how to use the GPU to render a 3D object
//it also shows how to do stereoscopic 3D
//it uses GS which is a WIP GPU abstraction layer that's currently part of 3DScraft
//keep in mind GPU reverse engineering is an ongoing effort and our understanding of it is still fairly limited.
#include <stdlib.h> #include <stdlib.h>
#include <stdio.h> #include <stdio.h>
#include <math.h> #include <math.h>
#include <string.h>
#include <3ds.h> #include <3ds.h>
#include "math.h" #include "math.h"
#include "gs.h"
#include "test_vsh_shbin.h" #include "test_vsh_shbin.h"
#include "test_png_bin.h" #include "texture_bin.h"
#include "mdl.h"
//will be moved into ctrulib at some point
#define CONFIG_3D_SLIDERSTATE (*(float*)0x1FF81080)
#define RGBA8(r,g,b,a) ((((r)&0xFF)<<24) | (((g)&0xFF)<<16) | (((b)&0xFF)<<8) | (((a)&0xFF)<<0))
//shader structure
DVLB_s* shader; DVLB_s* shader;
float* vertArray; //texture data pointer
u32* texData; u32* texData;
//vbo structure
gsVbo_s vbo;
void setUniformMatrix(u32 startreg, float* m) //GPU framebuffer address
{
float param[16];
param[0x0]=m[3]; //w
param[0x1]=m[2]; //z
param[0x2]=m[1]; //y
param[0x3]=m[0]; //x
param[0x4]=m[7];
param[0x5]=m[6];
param[0x6]=m[5];
param[0x7]=m[4];
param[0x8]=m[11];
param[0x9]=m[10];
param[0xa]=m[9];
param[0xb]=m[8];
param[0xc]=m[15];
param[0xd]=m[14];
param[0xe]=m[13];
param[0xf]=m[12];
GPU_SetUniform(startreg, (u32*)param, 4);
}
float angle=0.0f;
float angleZ=0.0f;
float tx, ty, tz;
u32* gpuOut=(u32*)0x1F119400; u32* gpuOut=(u32*)0x1F119400;
//GPU depth buffer address
u32* gpuDOut=(u32*)0x1F370800; u32* gpuDOut=(u32*)0x1F370800;
// topscreen //angle for the vertex lighting (cf test.vsh)
void doFrame1() float lightAngle;
//object position and rotation angle
vect3Df_s position, angle;
//vertex structure
typedef struct
{
vect3Df_s position;
float texcoord[2];
vect3Df_s normal;
}vertex_s;
//object data (cube)
//obviously this doesn't have to be defined manually, but we will here for the purposes of the example
//each line is a vertex : {position.x, position.y, position.z}, {texcoord.t, texcoord.s}, {normal.x, normal.y, normal.z}
//we're drawing triangles so three lines = one triangle
const vertex_s modelVboData[]=
{
//first face (PZ)
//first triangle
{(vect3Df_s){-0.5f, -0.5f, +0.5f}, (float[]){0.0f, 1.0f}, (vect3Df_s){0.0f, 0.0f, +1.0f}},
{(vect3Df_s){+0.5f, -0.5f, +0.5f}, (float[]){1.0f, 1.0f}, (vect3Df_s){0.0f, 0.0f, +1.0f}},
{(vect3Df_s){+0.5f, +0.5f, +0.5f}, (float[]){1.0f, 0.0f}, (vect3Df_s){0.0f, 0.0f, +1.0f}},
//second triangle
{(vect3Df_s){+0.5f, +0.5f, +0.5f}, (float[]){1.0f, 0.0f}, (vect3Df_s){0.0f, 0.0f, +1.0f}},
{(vect3Df_s){-0.5f, +0.5f, +0.5f}, (float[]){0.0f, 0.0f}, (vect3Df_s){0.0f, 0.0f, +1.0f}},
{(vect3Df_s){-0.5f, -0.5f, +0.5f}, (float[]){0.0f, 1.0f}, (vect3Df_s){0.0f, 0.0f, +1.0f}},
//second face (MZ)
//first triangle
{(vect3Df_s){-0.5f, -0.5f, -0.5f}, (float[]){0.0f, 1.0f}, (vect3Df_s){0.0f, 0.0f, -1.0f}},
{(vect3Df_s){-0.5f, +0.5f, -0.5f}, (float[]){1.0f, 1.0f}, (vect3Df_s){0.0f, 0.0f, -1.0f}},
{(vect3Df_s){+0.5f, +0.5f, -0.5f}, (float[]){1.0f, 0.0f}, (vect3Df_s){0.0f, 0.0f, -1.0f}},
//second triangle
{(vect3Df_s){+0.5f, +0.5f, -0.5f}, (float[]){1.0f, 0.0f}, (vect3Df_s){0.0f, 0.0f, -1.0f}},
{(vect3Df_s){+0.5f, -0.5f, -0.5f}, (float[]){0.0f, 0.0f}, (vect3Df_s){0.0f, 0.0f, -1.0f}},
{(vect3Df_s){-0.5f, -0.5f, -0.5f}, (float[]){0.0f, 1.0f}, (vect3Df_s){0.0f, 0.0f, -1.0f}},
//third face (PX)
//first triangle
{(vect3Df_s){+0.5f, -0.5f, -0.5f}, (float[]){0.0f, 1.0f}, (vect3Df_s){+1.0f, 0.0f, 0.0f}},
{(vect3Df_s){+0.5f, +0.5f, -0.5f}, (float[]){1.0f, 1.0f}, (vect3Df_s){+1.0f, 0.0f, 0.0f}},
{(vect3Df_s){+0.5f, +0.5f, +0.5f}, (float[]){1.0f, 0.0f}, (vect3Df_s){+1.0f, 0.0f, 0.0f}},
//second triangle
{(vect3Df_s){+0.5f, +0.5f, +0.5f}, (float[]){1.0f, 0.0f}, (vect3Df_s){+1.0f, 0.0f, 0.0f}},
{(vect3Df_s){+0.5f, -0.5f, +0.5f}, (float[]){0.0f, 0.0f}, (vect3Df_s){+1.0f, 0.0f, 0.0f}},
{(vect3Df_s){+0.5f, -0.5f, -0.5f}, (float[]){0.0f, 1.0f}, (vect3Df_s){+1.0f, 0.0f, 0.0f}},
//fourth face (MX)
//first triangle
{(vect3Df_s){-0.5f, -0.5f, -0.5f}, (float[]){0.0f, 1.0f}, (vect3Df_s){-1.0f, 0.0f, 0.0f}},
{(vect3Df_s){-0.5f, -0.5f, +0.5f}, (float[]){1.0f, 1.0f}, (vect3Df_s){-1.0f, 0.0f, 0.0f}},
{(vect3Df_s){-0.5f, +0.5f, +0.5f}, (float[]){1.0f, 0.0f}, (vect3Df_s){-1.0f, 0.0f, 0.0f}},
//second triangle
{(vect3Df_s){-0.5f, +0.5f, +0.5f}, (float[]){1.0f, 0.0f}, (vect3Df_s){-1.0f, 0.0f, 0.0f}},
{(vect3Df_s){-0.5f, +0.5f, -0.5f}, (float[]){0.0f, 0.0f}, (vect3Df_s){-1.0f, 0.0f, 0.0f}},
{(vect3Df_s){-0.5f, -0.5f, -0.5f}, (float[]){0.0f, 1.0f}, (vect3Df_s){-1.0f, 0.0f, 0.0f}},
//fifth face (PY)
//first triangle
{(vect3Df_s){-0.5f, +0.5f, -0.5f}, (float[]){0.0f, 1.0f}, (vect3Df_s){0.0f, +1.0f, 0.0f}},
{(vect3Df_s){-0.5f, +0.5f, +0.5f}, (float[]){1.0f, 1.0f}, (vect3Df_s){0.0f, +1.0f, 0.0f}},
{(vect3Df_s){+0.5f, +0.5f, +0.5f}, (float[]){1.0f, 0.0f}, (vect3Df_s){0.0f, +1.0f, 0.0f}},
//second triangle
{(vect3Df_s){+0.5f, +0.5f, +0.5f}, (float[]){1.0f, 0.0f}, (vect3Df_s){0.0f, +1.0f, 0.0f}},
{(vect3Df_s){+0.5f, +0.5f, -0.5f}, (float[]){0.0f, 0.0f}, (vect3Df_s){0.0f, +1.0f, 0.0f}},
{(vect3Df_s){-0.5f, +0.5f, -0.5f}, (float[]){0.0f, 1.0f}, (vect3Df_s){0.0f, +1.0f, 0.0f}},
//sixth face (MY)
//first triangle
{(vect3Df_s){-0.5f, -0.5f, -0.5f}, (float[]){0.0f, 1.0f}, (vect3Df_s){0.0f, -1.0f, 0.0f}},
{(vect3Df_s){+0.5f, -0.5f, -0.5f}, (float[]){1.0f, 1.0f}, (vect3Df_s){0.0f, -1.0f, 0.0f}},
{(vect3Df_s){+0.5f, -0.5f, +0.5f}, (float[]){1.0f, 0.0f}, (vect3Df_s){0.0f, -1.0f, 0.0f}},
//second triangle
{(vect3Df_s){+0.5f, -0.5f, +0.5f}, (float[]){1.0f, 0.0f}, (vect3Df_s){0.0f, -1.0f, 0.0f}},
{(vect3Df_s){-0.5f, -0.5f, +0.5f}, (float[]){0.0f, 0.0f}, (vect3Df_s){0.0f, -1.0f, 0.0f}},
{(vect3Df_s){-0.5f, -0.5f, -0.5f}, (float[]){0.0f, 1.0f}, (vect3Df_s){0.0f, -1.0f, 0.0f}},
};
//stolen from staplebutt
void GPU_SetDummyTexEnv(u8 num)
{
GPU_SetTexEnv(num,
GPU_TEVSOURCES(GPU_PREVIOUS, 0, 0),
GPU_TEVSOURCES(GPU_PREVIOUS, 0, 0),
GPU_TEVOPERANDS(0,0,0),
GPU_TEVOPERANDS(0,0,0),
GPU_REPLACE,
GPU_REPLACE,
0xFFFFFFFF);
}
// topscreen
void renderFrame()
{ {
//general setup
GPU_SetViewport((u32*)osConvertVirtToPhys((u32)gpuDOut),(u32*)osConvertVirtToPhys((u32)gpuOut),0,0,240*2,400); GPU_SetViewport((u32*)osConvertVirtToPhys((u32)gpuDOut),(u32*)osConvertVirtToPhys((u32)gpuOut),0,0,240*2,400);
GPU_DepthRange(-1.0f, 0.0f); GPU_DepthRange(-1.0f, 0.0f);
GPU_SetFaceCulling(GPU_CULL_BACK_CCW); GPU_SetFaceCulling(GPU_CULL_BACK_CCW);
GPU_SetStencilTest(false, GPU_ALWAYS, 0x00); GPU_SetStencilTest(false, GPU_ALWAYS, 0x00, 0xFF, 0x00);
GPU_SetStencilOp(GPU_KEEP, GPU_KEEP, GPU_KEEP);
GPU_SetBlendingColor(0,0,0,0);
GPU_SetDepthTestAndWriteMask(true, GPU_GREATER, GPU_WRITE_ALL); GPU_SetDepthTestAndWriteMask(true, GPU_GREATER, GPU_WRITE_ALL);
// ? GPUCMD_AddSingleParam(0x00010062, 0);
GPUCMD_AddSingleParam(0x00010062, 0x00000000); //param always 0x0 according to code GPUCMD_AddSingleParam(0x000F0118, 0);
GPUCMD_AddSingleParam(0x000F0118, 0x00000000);
//setup shader //setup shader
SHDR_UseProgram(shader, 0); SHDR_UseProgram(shader, 0);
//attribute buffers GPU_SetAlphaBlending(GPU_BLEND_ADD, GPU_BLEND_ADD, GPU_SRC_ALPHA, GPU_ONE_MINUS_SRC_ALPHA, GPU_SRC_ALPHA, GPU_ONE_MINUS_SRC_ALPHA);
GPU_SetAttributeBuffers(3, (u32*)osConvertVirtToPhys((u32)vertArray), GPU_SetAlphaTest(false, GPU_ALWAYS, 0x00);
GPU_SetTextureEnable(GPU_TEXUNIT0);
GPU_SetTexEnv(0,
GPU_TEVSOURCES(GPU_TEXTURE0, GPU_PRIMARY_COLOR, GPU_PRIMARY_COLOR),
GPU_TEVSOURCES(GPU_TEXTURE0, GPU_PRIMARY_COLOR, GPU_PRIMARY_COLOR),
GPU_TEVOPERANDS(0,0,0),
GPU_TEVOPERANDS(0,0,0),
GPU_MODULATE, GPU_MODULATE,
0xFFFFFFFF);
GPU_SetDummyTexEnv(1);
GPU_SetDummyTexEnv(2);
GPU_SetDummyTexEnv(3);
GPU_SetDummyTexEnv(4);
GPU_SetDummyTexEnv(5);
//texturing stuff
GPU_SetTexture(GPU_TEXUNIT0, (u32*)osConvertVirtToPhys((u32)texData),128,128,GPU_TEXTURE_MAG_FILTER(GPU_NEAREST)|GPU_TEXTURE_MIN_FILTER(GPU_NEAREST),GPU_RGBA8);
GPU_SetAttributeBuffers(3, (u32*)osConvertVirtToPhys((u32)texData),
GPU_ATTRIBFMT(0, 3, GPU_FLOAT)|GPU_ATTRIBFMT(1, 2, GPU_FLOAT)|GPU_ATTRIBFMT(2, 3, GPU_FLOAT), GPU_ATTRIBFMT(0, 3, GPU_FLOAT)|GPU_ATTRIBFMT(1, 2, GPU_FLOAT)|GPU_ATTRIBFMT(2, 3, GPU_FLOAT),
0xFFC, 0x210, 1, (u32[]){0x00000000}, (u64[]){0x210}, (u8[]){3}); 0xFFC, 0x210, 1, (u32[]){0x00000000}, (u64[]){0x210}, (u8[]){3});
//? //setup lighting (this is specific to our shader)
GPUCMD_AddSingleParam(0x000F0100, 0x00E40100); vect3Df_s lightDir=vnormf(vect3Df(cos(lightAngle), -1.0f, sin(lightAngle)));
GPUCMD_AddSingleParam(0x000F0101, 0x01010000); GPU_SetUniform(SHDR_GetUniformRegister(shader, "lightDirection", 0), (u32*)(float[]){0.0f, -lightDir.z, -lightDir.y, -lightDir.x}, 4);
GPUCMD_AddSingleParam(0x000F0104, 0x00000010); GPU_SetUniform(SHDR_GetUniformRegister(shader, "lightAmbient", 0), (u32*)(float[]){0.7f, 0.4f, 0.4f, 0.4f}, 4);
//texturing stuff //initialize projection matrix to standard perspective stuff
GPUCMD_AddSingleParam(0x0002006F, 0x00000100); gsMatrixMode(GS_PROJECTION);
GPUCMD_AddSingleParam(0x000F0080, 0x00011001); //enables/disables texturing gsProjectionMatrix(80.0f*M_PI/180.0f, 240.0f/400.0f, 0.01f, 100.0f);
gsRotateZ(M_PI/2); //because framebuffer is sideways...
//texenv //draw object
GPU_SetTexEnv(3, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00000000); gsMatrixMode(GS_MODELVIEW);
GPU_SetTexEnv(4, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00000000); gsPushMatrix();
GPU_SetTexEnv(5, GPU_TEVSOURCES(GPU_TEXTURE0, GPU_PRIMARY_COLOR, GPU_PRIMARY_COLOR), GPU_TEVSOURCES(GPU_TEXTURE0, GPU_PRIMARY_COLOR, GPU_PRIMARY_COLOR), gsTranslate(position.x, position.y, position.z);
GPU_TEVOPERANDS(0,0,0), GPU_TEVOPERANDS(0,0,0), GPU_MODULATE, GPU_MODULATE, 0xFFFFFFFF); gsRotateX(angle.x);
gsRotateY(angle.y);
//texturing stuff gsVboDraw(&vbo);
GPU_SetTexture((u32*)osConvertVirtToPhys((u32)texData),256,256,0x6,GPU_RGBA8); gsPopMatrix();
//setup matrices
float modelView[16];
float projection[16];
loadIdentity44(modelView);
loadIdentity44(projection);
translateMatrix(modelView, tx, ty, tz);
rotateMatrixX(modelView, angle);
rotateMatrixZ(modelView, angleZ);
initProjectionMatrix(projection, 1.3962634f, 240.0f/400.0f, 0.01f, 10.0f);
setUniformMatrix(0x24, modelView);
setUniformMatrix(0x20, projection);
//draw first model
GPU_DrawArray(GPU_TRIANGLES, mdlFaces*3);
// GPU_DrawElements(GPU_TRIANGLES, (u32*)(((u32)((void*)indArray-(void*)gspHeap))+0x20000000-base), 6);
//setup matrices
loadIdentity44(modelView);
loadIdentity44(projection);
translateMatrix(modelView, tx, -ty, tz);
rotateMatrixX(modelView, -angle);
rotateMatrixZ(modelView, -angleZ);
setUniformMatrix(0x24, modelView);
//draw second
GPU_DrawArray(GPU_TRIANGLES, mdlFaces*3);
//finalize stuff ?
GPU_FinishDrawing(); GPU_FinishDrawing();
} }
void demoControls(void) int main(int argc, char** argv)
{
hidScanInput();
u32 PAD=hidKeysHeld();
if(PAD&KEY_UP)tx+=0.1f;
if(PAD&KEY_DOWN)tx-=0.1f;
if(PAD&KEY_LEFT)ty+=0.1f;
if(PAD&KEY_RIGHT)ty-=0.1f;
if(PAD&KEY_R)tz+=0.1f;
if(PAD&KEY_L)tz-=0.1f;
if(PAD&KEY_A)angle+=0.1f;
if(PAD&KEY_Y)angle-=0.1f;
if(PAD&KEY_X)angleZ+=0.1f;
if(PAD&KEY_B)angleZ-=0.1f;
}
extern u32* gxCmdBuf;
int main()
{ {
//setup services
srvInit(); srvInit();
aptInit(); aptInit();
gfxInit(); gfxInit();
hidInit(NULL); hidInit(NULL);
//initialize GPU
GPU_Init(NULL); GPU_Init(NULL);
//let GFX know we're ok with doing stereoscopic 3D rendering
gfxSet3D(true);
//load our vertex shader binary
shader=SHDR_ParseSHBIN((u32*)test_vsh_shbin, test_vsh_shbin_size);
//initialize GS
gsInit(shader);
//allocate our GPU command buffers
//they *have* to be on the linear heap
u32 gpuCmdSize=0x40000; u32 gpuCmdSize=0x40000;
u32* gpuCmd=(u32*)linearAlloc(gpuCmdSize*4); u32* gpuCmd=(u32*)linearAlloc(gpuCmdSize*4);
u32* gpuCmdRight=(u32*)linearAlloc(gpuCmdSize*4);
GPU_Reset(gxCmdBuf, gpuCmd, gpuCmdSize); //actually reset the GPU
GPU_Reset(NULL, gpuCmd, gpuCmdSize);
vertArray=(float*)linearAlloc(0x100000); //create texture
texData=(u32*)linearAlloc(0x100000); texData=(u32*)linearMemAlign(texture_bin_size, 0x80); //textures need to be 0x80-byte aligned
memcpy(texData, texture_bin, texture_bin_size);
memcpy(texData, test_png_bin, test_png_bin_size); //create VBO
memcpy(vertArray, mdlData, sizeof(mdlData)); gsVboInit(&vbo);
GSPGPU_FlushDataCache(NULL, mdlData, sizeof(mdlData)); gsVboCreate(&vbo, sizeof(modelVboData));
GSPGPU_FlushDataCache(NULL, test_png_bin, test_png_bin_size); gsVboAddData(&vbo, (void*)modelVboData, sizeof(modelVboData), sizeof(modelVboData)/sizeof(vertex_s));
gsVboFlushData(&vbo);
tx=ty=0.0f; tz=-0.1f; //initialize object position and angle
shader=SHDR_ParseSHBIN((u32*)test_vsh_shbin,test_vsh_shbin_size); position=vect3Df(0.0f, 0.0f, -2.0f);
angle=vect3Df(M_PI/4, M_PI/4, 0.0f);
GX_SetMemoryFill(gxCmdBuf, (u32*)gpuOut, 0x404040FF, (u32*)&gpuOut[0x2EE00], 0x201, (u32*)gpuDOut, 0x00000000, (u32*)&gpuDOut[0x2EE00], 0x201); //background color (blue)
gspWaitForPSC0(); u32 backgroundColor=RGBA8(0x68, 0xB0, 0xD8, 0xFF);
gfxSwapBuffersGpu();
while(aptMainLoop()) while(aptMainLoop())
{ {
demoControls(); //get current 3D slider state
float slider=CONFIG_3D_SLIDERSTATE;
GX_SetMemoryFill(gxCmdBuf, (u32*)gpuOut, 0x404040FF, (u32*)&gpuOut[0x2EE00], 0x201, (u32*)gpuDOut, 0x00000000, (u32*)&gpuDOut[0x2EE00], 0x201); //controls
hidScanInput();
//START to exit to hbmenu
if(keysDown()&KEY_START)break;
//A/B to change vertex lighting angle
if(keysHeld()&KEY_A)lightAngle+=0.1f;
if(keysHeld()&KEY_B)lightAngle-=0.1f;
//D-PAD to rotate object
if(keysHeld()&KEY_RIGHT)angle.x+=0.05f;
if(keysHeld()&KEY_LEFT)angle.x-=0.05f;
if(keysHeld()&KEY_UP)angle.y+=0.05f;
if(keysHeld()&KEY_DOWN)angle.y-=0.05f;
//R/L to bring object closer to or move it further from the camera
if(keysHeld()&KEY_R)position.z+=0.1f;
if(keysHeld()&KEY_L)position.z-=0.1f;
//generate our GPU command buffer for this frame
gsStartFrame();
renderFrame();
GPUCMD_Finalize();
if(slider>0.0f)
{
//new and exciting 3D !
//make a copy of left gpu buffer
u32 offset; GPUCMD_GetBuffer(NULL, NULL, &offset);
memcpy(gpuCmdRight, gpuCmd, offset*4);
//setup interaxial
float interaxial=slider*0.12f;
//adjust left gpu buffer fo 3D !
{mtx44 m; loadIdentity44((float*)m); translateMatrix((float*)m, -interaxial*0.5f, 0.0f, 0.0f); gsAdjustBufferMatrices(m);}
//draw left framebuffer
GPUCMD_FlushAndRun(NULL);
//while GPU starts drawing the left buffer, adjust right one for 3D !
GPUCMD_SetBuffer(gpuCmdRight, gpuCmdSize, offset);
{mtx44 m; loadIdentity44((float*)m); translateMatrix((float*)m, interaxial*0.5f, 0.0f, 0.0f); gsAdjustBufferMatrices(m);}
//we wait for the left buffer to finish drawing
gspWaitForP3D();
GX_SetDisplayTransfer(NULL, (u32*)gpuOut, 0x019001E0, (u32*)gfxGetFramebuffer(GFX_TOP, GFX_LEFT, NULL, NULL), 0x019001E0, 0x01001000);
gspWaitForPPF();
//we draw the right buffer, wait for it to finish and then switch back to left one
//clear the screen
GX_SetMemoryFill(NULL, (u32*)gpuOut, backgroundColor, (u32*)&gpuOut[0x2EE00], 0x201, (u32*)gpuDOut, 0x00000000, (u32*)&gpuDOut[0x2EE00], 0x201);
gspWaitForPSC0(); gspWaitForPSC0();
GPUCMD_SetBuffer(gpuCmd, gpuCmdSize, 0); //draw the right framebuffer
doFrame1(); GPUCMD_FlushAndRun(NULL);
GPUCMD_Finalize();
GPUCMD_Run(gxCmdBuf);
gspWaitForP3D(); gspWaitForP3D();
gfxSwapBuffersGpu(); //transfer from GPU output buffer to actual framebuffer
GX_SetDisplayTransfer(gxCmdBuf, (u32*)gpuOut, 0x019001E0, (u32*)gfxGetFramebuffer(GFX_TOP, GFX_LEFT, NULL, NULL), 0x019001E0, 0x01001000); GX_SetDisplayTransfer(NULL, (u32*)gpuOut, 0x019001E0, (u32*)gfxGetFramebuffer(GFX_TOP, GFX_RIGHT, NULL, NULL), 0x019001E0, 0x01001000);
gspWaitForPPF();
GPUCMD_SetBuffer(gpuCmd, gpuCmdSize, 0);
}else{
//boring old 2D !
//draw the frame
GPUCMD_FlushAndRun(NULL);
gspWaitForP3D();
//clear the screen
GX_SetDisplayTransfer(NULL, (u32*)gpuOut, 0x019001E0, (u32*)gfxGetFramebuffer(GFX_TOP, GFX_LEFT, NULL, NULL), 0x019001E0, 0x01001000);
gspWaitForPPF(); gspWaitForPPF();
gspWaitForVBlank();
} }
//clear the screen
GX_SetMemoryFill(NULL, (u32*)gpuOut, backgroundColor, (u32*)&gpuOut[0x2EE00], 0x201, (u32*)gpuDOut, 0x00000000, (u32*)&gpuDOut[0x2EE00], 0x201);
gspWaitForPSC0();
gfxSwapBuffersGpu();
gspWaitForEvent(GSPEVENT_VBlank0, true);
}
gsExit();
hidExit(); hidExit();
gfxExit(); gfxExit();
aptExit(); aptExit();

101
examples/gpu/source/math.c
View File

@ -15,6 +15,7 @@ void multMatrix44(float* m1, float* m2, float* m) //4x4
{ {
int i, j; int i, j;
for(i=0;i<4;i++)for(j=0;j<4;j++)m[i+j*4]=(m1[0+j*4]*m2[i+0*4])+(m1[1+j*4]*m2[i+1*4])+(m1[2+j*4]*m2[i+2*4])+(m1[3+j*4]*m2[i+3*4]); for(i=0;i<4;i++)for(j=0;j<4;j++)m[i+j*4]=(m1[0+j*4]*m2[i+0*4])+(m1[1+j*4]*m2[i+1*4])+(m1[2+j*4]*m2[i+2*4])+(m1[3+j*4]*m2[i+3*4]);
} }
void translateMatrix(float* tm, float x, float y, float z) void translateMatrix(float* tm, float x, float y, float z)
@ -26,11 +27,16 @@ void translateMatrix(float* tm, float x, float y, float z)
rm[7]=y; rm[7]=y;
rm[11]=z; rm[11]=z;
multMatrix44(rm,tm,m); multMatrix44(tm,rm,m);
memcpy(tm,m,16*sizeof(float)); memcpy(tm,m,16*sizeof(float));
} }
void rotateMatrixX(float* tm, float x) // 00 01 02 03
// 04 05 06 07
// 08 09 10 11
// 12 13 14 15
void rotateMatrixX(float* tm, float x, bool r)
{ {
float rm[16], m[16]; float rm[16], m[16];
memset(rm, 0x00, 16*4); memset(rm, 0x00, 16*4);
@ -40,11 +46,27 @@ void rotateMatrixX(float* tm, float x)
rm[9]=-sin(x); rm[9]=-sin(x);
rm[10]=cos(x); rm[10]=cos(x);
rm[15]=1.0f; rm[15]=1.0f;
multMatrix44(tm,rm,m); if(!r)multMatrix44(tm,rm,m);
else multMatrix44(rm,tm,m);
memcpy(tm,m,16*sizeof(float)); memcpy(tm,m,16*sizeof(float));
} }
void rotateMatrixZ(float* tm, float x) void rotateMatrixY(float* tm, float x, bool r)
{
float rm[16], m[16];
memset(rm, 0x00, 16*4);
rm[0]=cos(x);
rm[2]=sin(x);
rm[5]=1.0f;
rm[8]=-sin(x);
rm[10]=cos(x);
rm[15]=1.0f;
if(!r)multMatrix44(tm,rm,m);
else multMatrix44(rm,tm,m);
memcpy(tm,m,16*sizeof(float));
}
void rotateMatrixZ(float* tm, float x, bool r)
{ {
float rm[16], m[16]; float rm[16], m[16];
memset(rm, 0x00, 16*4); memset(rm, 0x00, 16*4);
@ -54,7 +76,8 @@ void rotateMatrixZ(float* tm, float x)
rm[5]=cos(x); rm[5]=cos(x);
rm[10]=1.0f; rm[10]=1.0f;
rm[15]=1.0f; rm[15]=1.0f;
multMatrix44(tm,rm,m); if(!r)multMatrix44(tm,rm,m);
else multMatrix44(rm,tm,m);
memcpy(tm,m,16*sizeof(float)); memcpy(tm,m,16*sizeof(float));
} }
@ -70,26 +93,56 @@ void initProjectionMatrix(float* m, float fovy, float aspect, float near, float
float top = near*tan(fovy/2); float top = near*tan(fovy/2);
float right = (top*aspect); float right = (top*aspect);
*(m++) = near/right; float mp[4*4];
*(m++) = 0.0f;
*(m++) = 0.0f;
*(m++) = 0.0f;
*(m++) = 0.0f; mp[0x0] = near/right;
*(m++) = near/top; mp[0x1] = 0.0f;
*(m++) = 0.0f; mp[0x2] = 0.0f;
*(m++) = 0.0f; mp[0x3] = 0.0f;
*(m++) = 0.0f; mp[0x4] = 0.0f;
*(m++) = 0.0f; mp[0x5] = near/top;
// *(m++) = -(far+near)/(far-near); mp[0x6] = 0.0f;
*(m++) = 0.0f; mp[0x7] = 0.0f;
// *(m++) = -2.0f*(far*near)/(far-near);
// *(m++) = 1.0f;
*(m++) = -1.0f;
*(m++) = 0.0f; mp[0x8] = 0.0f;
*(m++) = 0.0f; mp[0x9] = 0.0f;
*(m++) = -1.0f; mp[0xA] = -(far+near)/(far-near);
*(m++) = 0.0f; mp[0xB] = -2.0f*(far*near)/(far-near);
mp[0xC] = 0.0f;
mp[0xD] = 0.0f;
mp[0xE] = -1.0f;
mp[0xF] = 0.0f;
float mp2[4*4];
loadIdentity44(mp2);
mp2[0xA]=0.5;
mp2[0xB]=-0.5;
multMatrix44(mp2, mp, m);
}
vect3Df_s getMatrixColumn(float* m, u8 i)
{
if(!m || i>=4)return vect3Df(0,0,0);
return vect3Df(m[0+i*4],m[1+i*4],m[2+i*4]);
}
vect3Df_s getMatrixRow(float* m, u8 i)
{
if(!m || i>=4)return vect3Df(0,0,0);
return vect3Df(m[i+0*4],m[i+1*4],m[i+2*4]);
}
vect4Df_s getMatrixColumn4(float* m, u8 i)
{
if(!m || i>=4)return vect4Df(0,0,0,0);
return vect4Df(m[0+i*4],m[1+i*4],m[2+i*4],m[3+i*4]);
}
vect4Df_s getMatrixRow4(float* m, u8 i)
{
if(!m || i>=4)return vect4Df(0,0,0,0);
return vect4Df(m[i+0*4],m[i+1*4],m[i+2*4],m[i+3*4]);
} }

135
examples/gpu/source/math.h
View File

@ -1,13 +1,144 @@
#ifndef MATH_H #ifndef MATH_H
#define MATH_H
#include <3ds/types.h>
#include <math.h>
typedef float mtx44[4][4];
typedef float mtx33[3][3];
typedef struct
{
s32 x, y, z;
}vect3Di_s;
static inline vect3Di_s vect3Di(s32 x, s32 y, s32 z)
{
return (vect3Di_s){x,y,z};
}
static inline vect3Di_s vaddi(vect3Di_s u, vect3Di_s v)
{
return (vect3Di_s){u.x+v.x,u.y+v.y,u.z+v.z};
}
static inline vect3Di_s vsubi(vect3Di_s u, vect3Di_s v)
{
return (vect3Di_s){u.x-v.x,u.y-v.y,u.z-v.z};
}
static inline vect3Di_s vmuli(vect3Di_s v, s32 f)
{
return (vect3Di_s){v.x*f,v.y*f,v.z*f};
}
typedef struct
{
float x, y, z;
}vect3Df_s;
static inline vect3Df_s vect3Df(float x, float y, float z)
{
return (vect3Df_s){x,y,z};
}
static inline vect3Df_s vaddf(vect3Df_s u, vect3Df_s v)
{
return (vect3Df_s){u.x+v.x,u.y+v.y,u.z+v.z};
}
static inline vect3Df_s vsubf(vect3Df_s u, vect3Df_s v)
{
return (vect3Df_s){u.x-v.x,u.y-v.y,u.z-v.z};
}
static inline vect3Df_s vmulf(vect3Df_s v, float f)
{
return (vect3Df_s){v.x*f,v.y*f,v.z*f};
}
static inline vect3Df_s vscalef(vect3Df_s v1, vect3Df_s v2)
{
return (vect3Df_s){v1.x*v2.x,v1.y*v2.y,v1.z*v2.z};
}
static inline float vmagf(vect3Df_s v)
{
return sqrtf(v.x*v.x+v.y*v.y+v.z*v.z);
}
static inline float vdistf(vect3Df_s v1, vect3Df_s v2)
{
return sqrtf((v1.x-v2.x)*(v1.x-v2.x)+(v1.y-v2.y)*(v1.y-v2.y)+(v1.z-v2.z)*(v1.z-v2.z));
}
static inline vect3Df_s vnormf(vect3Df_s v)
{
const float l=sqrtf(v.x*v.x+v.y*v.y+v.z*v.z);
return (vect3Df_s){v.x/l,v.y/l,v.z/l};
}
typedef struct
{
float x, y, z, w;
}vect4Df_s;
static inline vect4Df_s vect4Df(float x, float y, float z, float w)
{
return (vect4Df_s){x,y,z,w};
}
static inline vect4Df_s vaddf4(vect4Df_s u, vect4Df_s v)
{
return (vect4Df_s){u.x+v.x,u.y+v.y,u.z+v.z,u.w+v.w};
}
static inline vect4Df_s vsubf4(vect4Df_s u, vect4Df_s v)
{
return (vect4Df_s){u.x-v.x,u.y-v.y,u.z-v.z,u.w-v.w};
}
static inline vect4Df_s vmulf4(vect4Df_s v, float f)
{
return (vect4Df_s){v.x*f,v.y*f,v.z*f,v.w*f};
}
static inline float vdotf4(vect4Df_s v1, vect4Df_s v2)
{
return v1.x*v2.x+v1.y*v2.y+v1.z*v2.z+v1.w*v2.w;
}
static inline vect4Df_s vnormf4(vect4Df_s v)
{
const float l=sqrtf(v.x*v.x+v.y*v.y+v.z*v.z+v.w*v.w);
return (vect4Df_s){v.x/l,v.y/l,v.z/l,v.w/l};
}
//interstuff
static inline vect3Di_s vf2i(vect3Df_s v)
{
return (vect3Di_s){floorf(v.x),floorf(v.y),floorf(v.z)};
}
static inline vect3Df_s vi2f(vect3Di_s v)
{
return (vect3Df_s){(float)v.x,(float)v.y,(float)v.z};
}
void loadIdentity44(float* m); void loadIdentity44(float* m);
void multMatrix44(float* m1, float* m2, float* m); void multMatrix44(float* m1, float* m2, float* m);
void translateMatrix(float* tm, float x, float y, float z); void translateMatrix(float* tm, float x, float y, float z);
void rotateMatrixX(float* tm, float x); void rotateMatrixX(float* tm, float x, bool r);
void rotateMatrixZ(float* tm, float x); void rotateMatrixY(float* tm, float x, bool r);
void rotateMatrixZ(float* tm, float x, bool r);
void scaleMatrix(float* tm, float x, float y, float z); void scaleMatrix(float* tm, float x, float y, float z);
void initProjectionMatrix(float* m, float fovy, float aspect, float near, float far); void initProjectionMatrix(float* m, float fovy, float aspect, float near, float far);
vect3Df_s getMatrixColumn(float* m, u8 i);
vect3Df_s getMatrixRow(float* m, u8 i);
vect4Df_s getMatrixColumn4(float* m, u8 i);
vect4Df_s getMatrixRow4(float* m, u8 i);
#endif #endif