+test

17 files changed, 1229 insertions, 0 deletions

diff --git a/src/core/clip.c b/src/core/clip.c
new file mode 100644
index 0000000..8fe1eeb
--- /dev/null
+++ b/src/core/clip.c
@@ -0,0 +1,201 @@
+#include "clip.h"
+#include "shader.h"
+
+typedef enum {
+	POSITIVE_W,
+	POSITIVE_X,
+	NEGATIVE_X,
+	POSITIVE_Y,
+	NEGATIVE_Y,
+	POSITIVE_Z,
+	NEGATIVE_Z
+} Plane;
+
+static bool is_inside_plane(Vec4* c, Plane plane) {
+	switch (plane) {
+		case POSITIVE_W:
+			return c->w >= EPSILON;
+		case POSITIVE_X:
+			return c->x <= +c->w;
+		case NEGATIVE_X:
+			return c->x >= -c->w;
+		case POSITIVE_Y:
+			return c->y <= +c->w;
+		case NEGATIVE_Y:
+			return c->y >= -c->w;
+		case POSITIVE_Z:
+			return c->z <= +c->w;
+		case NEGATIVE_Z:
+			return c->z >= -c->w;
+		default:
+			ssr_assert(0);
+			return 0;
+	}
+}
+
+static bool is_vertex_visible(Vec4* v) {
+	return absf(v->x) <= v->w
+		&& absf(v->y) <= v->w
+		&& absf(v->z) <= v->w;
+}
+
+static float get_intersect_ratio(Vec4* prev, Vec4* curr, Plane plane) {
+	switch (plane) {
+	case POSITIVE_W:
+		return (prev->w - EPSILON) / (prev->w - curr->w);
+	case POSITIVE_X:
+		return (prev->w - prev->x) / ((prev->w - prev->x) - (curr->w - curr->x));
+	case NEGATIVE_X:
+		return (prev->w + prev->x) / ((prev->w + prev->x) - (curr->w + curr->x));
+	case POSITIVE_Y:
+		return (prev->w - prev->y) / ((prev->w - prev->y) - (curr->w - curr->y));
+	case NEGATIVE_Y:
+		return (prev->w + prev->y) / ((prev->w + prev->y) - (curr->w + curr->y));
+	case POSITIVE_Z:
+		return (prev->w - prev->z) / ((prev->w - prev->z) - (curr->w - curr->z));
+	case NEGATIVE_Z:
+		return (prev->w + prev->z) / ((prev->w + prev->z) - (curr->w + curr->z));
+	default:
+		assert(0);
+		return 0;
+	}
+}
+
+static bool clip_against_plane(Plane plane, uint varying_flag, ClippedBuffer* src_buffer, ClippedBuffer* dst_buffer) {
+	int idx = 0;
+	for (int i = 0; i < src_buffer->count; ++i) {
+		ClippedVert* prev = &src_buffer->vertices[(i - 1 + src_buffer->count) % src_buffer->count];
+		ClippedVert* curr = &src_buffer->vertices[i];
+
+		bool prev_inside = is_inside_plane(&prev->clip_coord, plane);
+		bool curr_inside = is_inside_plane(&curr->clip_coord, plane);
+
+		if (prev_inside != curr_inside) {
+			ClippedVert* dst = &dst_buffer->vertices[idx];
+			float t = get_intersect_ratio(&prev->clip_coord, &curr->clip_coord, plane);
+			ssrS_lerpvec4(t, &prev->clip_coord, &curr->clip_coord, &dst->clip_coord);
+			/*set varying varibles*/
+			if (varying_flag & VARYING_BASIC) {
+				if (varying_flag & VARYING_POSITION) ssrS_lerpvec3(t, &prev->vertex.position, &curr->vertex.position, &dst->vertex.position);
+				if (varying_flag & VARYING_NORMAL) ssrS_lerpvec3(t, &prev->vertex.normal, &curr->vertex.normal, &dst->vertex.normal);
+				if (varying_flag & VARYING_TANGENT) ssrS_lerpvec3(t, &prev->vertex.tangent, &curr->vertex.tangent, &dst->vertex.tangent);
+				if (varying_flag & VARYING_TEXCOORD) ssrS_lerpvec2(t, &prev->vertex.texcoord, &curr->vertex.texcoord, &dst->vertex.texcoord);
+				if (varying_flag & VARYING_COLOR) ssrS_lerpcolor(t, prev->vertex.color, curr->vertex.color, &dst->vertex.color);
+				if (varying_flag & VARYING_JOINT) ssrS_lerpvec4(t, &prev->vertex.joint, &curr->vertex.joint, &dst->vertex.joint);
+				if (varying_flag & VARYING_WEIGHT) ssrS_lerpvec4(t, &prev->vertex.weight, &curr->vertex.weight, &dst->vertex.weight);
+			}
+			if (varying_flag & VARYING_EXTRA) {
+				int j = 0;
+				if (varying_flag & VARYING_NUM) {
+					for (j = 0; j < REG_NUM_COUNT; ++j) {
+						if (varying_flag & (VARYING_NUM_00 << j)) {
+							ssrS_lerpnum(t, prev->num[j], curr->num[j], &dst->num[j]);
+						}
+					}
+				}
+				if (varying_flag & VARYING_V2) {
+					for (j = 0; j < REG_V2_COUNT; ++j) {
+						if (varying_flag & (VARYING_V2_00 << j)) {
+							ssrS_lerpvec2(t, &prev->v2[j], &curr->v2[j], &dst->v2[j]);
+						}
+					}
+				}
+				if (varying_flag & VARYING_V3) {
+					for (j = 0; j < REG_V3_COUNT; ++j) {
+						if (varying_flag & (VARYING_V3_00 << j)) {
+							ssrS_lerpvec3(t, &prev->v3[j], &curr->v3[j], &dst->v3[j]);
+						}
+					}
+				}
+				if (varying_flag & VARYING_V4) {
+					for (j = 0; j < REG_V4_COUNT; ++j) {
+						if (varying_flag & (VARYING_V4_00 << j)) {
+							ssrS_lerpvec4(t, &prev->v4[j], &curr->v4[j], &dst->v4[j]);
+						}
+					}
+				}
+			}
+			++idx;
+		}
+
+		if (curr_inside) {
+			ClippedVert* dst = &dst_buffer->vertices[idx];
+			*dst = *curr;
+			++idx;
+		}
+	}
+	dst_buffer->count = idx;
+	return idx < 3;
+}
+
+#define CLIP(plane, from, to) \
+do {																					               \
+is_cull = clip_against_plane(plane, varying_flag, from, to); \
+if (is_cull) {																               \
+	buffer->count = 0;/*cull this triangle*/		               \
+	return 1;																		               \
+}																							               \
+}while(0)
+
+bool clip_triangle(Vec4* c0, Vec4* c1, Vec4* c2, Vert* v0, Vert* v1, Vert* v2, uint varying_flag, ClippedBuffer* buffer) {
+	bool is_visible = is_vertex_visible(c0)
+	&& is_vertex_visible(c1)
+	&& is_vertex_visible(c2);
+
+	if (is_visible) {
+		return 0; /*no need to clip*/
+	}
+
+	/*clipping it*/
+
+	static ClippedBuffer temp;
+
+	/*copy vert data to temp*/
+	temp.count = 3; 
+#define INIT_CLIP_VERT(idx) \
+	temp.vertices[idx].clip_coord = *c##idx;												 	 \
+	temp.vertices[idx].vertex = *v##idx;														 	 \
+	if (varying_flag & VARYING_EXTRA) {															 	 \
+		int i = 0;																										 	 \
+		if(varying_flag && VARYING_NUM) {															 	 \
+			for (i = 0; i < REG_NUM_COUNT; ++i) {													 \
+				if (varying_flag & (VARYING_NUM_00 << i))										 \
+					temp.vertices[idx].num[i] = reg_num[i].num[v##idx->index]; \
+			}																															 \
+		}																															 	 \
+		if(varying_flag && VARYING_V2) {															 	 \
+			for (i = 0; i < REG_V2_COUNT; ++i) {													 \
+				if (varying_flag & (VARYING_V2_00 << i))										 \
+					temp.vertices[idx].v2[i] = reg_v2[i].v2[v##idx->index];		 \
+			}																															 \
+		}																																 \
+		if(varying_flag && VARYING_V3) {															 	 \
+			for (i = 0; i < REG_V3_COUNT; ++i) {													 \
+				if (varying_flag & (VARYING_V3_00 << i))										 \
+					temp.vertices[idx].v3[i] = reg_v3[i].v3[v##idx->index];		 \
+			}																															 \
+		}																																 \
+		if(varying_flag && VARYING_V4) {															 	 \
+			for (i = 0; i < REG_V4_COUNT; ++i) {													 \
+				if (varying_flag & (VARYING_V4_00 << i))										 \
+					temp.vertices[idx].v4[i] = reg_v4[i].v4[v##idx->index];		 \
+			}																															 \
+		}																																 \
+	}
+	INIT_CLIP_VERT(0);
+	INIT_CLIP_VERT(1);
+	INIT_CLIP_VERT(2);
+#undef INIT_CLIP_VERT
+
+	bool is_cull = FALSE;
+
+	CLIP(POSITIVE_W, &temp, buffer);
+	CLIP(POSITIVE_X, buffer, &temp);
+	CLIP(NEGATIVE_X, &temp, buffer);
+	CLIP(POSITIVE_Y, buffer, &temp);
+	CLIP(NEGATIVE_Y, &temp, buffer);
+	CLIP(POSITIVE_Z, buffer, &temp);
+	CLIP(NEGATIVE_Z, &temp, buffer);
+
+	return 1;
+}
diff --git a/src/core/clip.h b/src/core/clip.h
new file mode 100644
index 0000000..b6f2702
--- /dev/null
+++ b/src/core/clip.h
@@ -0,0 +1,34 @@
+#ifndef _SOFTSHADEROOM_CLIP_H_
+#define _SOFTSHADEROOM_CLIP_H_
+
+#include "../util/type.h"
+#include "vert.h"
+
+#define LERP(t,a,b) ((1-(t))*(a)+(t)*(b))
+
+typedef struct ClippedVert {
+	/*clipping coord*/
+	Vec4 clip_coord;
+	/*vertex data*/
+	Vert vertex; 
+	/*register values*/
+	float num[4];
+	Vec2 v2[6];
+	Vec3 v3[8];
+	Vec4 v4[6];
+} ClippedVert;
+
+#define CLIP_BUFFER_SIZE 6
+
+typedef struct ClippedBuffer {
+	ClippedVert vertices[CLIP_BUFFER_SIZE];
+	uint count;
+} ClippedBuffer;
+
+bool clip_triangle(Vec4* c0, Vec4* c1, Vec4* c2, Vert* v0, Vert* v1, Vert* v2, uint varying_flag, ClippedBuffer* clipped);
+
+uint clip_line();
+
+uint clip_point();
+
+#endif
\ No newline at end of file
diff --git a/src/core/texture.c b/src/core/texture.c
new file mode 100644
index 0000000..7104bfb
--- /dev/null
+++ b/src/core/texture.c
@@ -0,0 +1,3 @@
+#include "texture.h"
+
+
diff --git a/src/core/texture.h b/src/core/texture.h
new file mode 100644
index 0000000..d6aee82
--- /dev/null
+++ b/src/core/texture.h
@@ -0,0 +1,11 @@
+#ifndef _SOFTSHADEROOM_TEXTURE_H_
+#define _SOFTSHADEROOM_TEXTURE_H_
+
+#include "vert.h"
+
+typedef struct Texture{
+	uint id;/*identifier*/
+	uint width, height;
+}Texture;
+
+#endif
\ No newline at end of file
diff --git a/src/example/01_dot.c b/src/example/01_dot.c
new file mode 100644
index 0000000..0a88084
--- /dev/null
+++ b/src/example/01_dot.c
@@ -0,0 +1,36 @@
+#include "example.h"
+#include "../core/rasterizer.h"
+
+Vec3 pos = { 0, 0, 2 }, target = { 0,0,-1 }, up = { 0, 1, 0 };
+
+float dot[] = { 0, 0, -2 };
+
+void onloaddot(void* data) {
+
+}
+
+void oneventdot(void* data) {
+	SDL_Event* e = (SDL_Event*)data;
+}
+
+void onupdatedot(void*data) {
+	uint dt = *(uint*)data;
+	ssr_matrixmode(MATRIX_VIEW);
+	ssr_lookat(&pos, &target, &up);
+	ssr_matrixmode(MATRIX_PROJECTION);
+}
+
+float j = 0;
+
+void ondrawdot(void*data) {
+	ssr_clearcolor(0xffffffff);
+
+	//for (int i = -50; i < 200; ++i) {
+	//	ssr_putpoint(i * sin(j += 0.1f), 300 * cos(j), ssr_color(0xff, 0, 0xff, 0));
+	//}
+
+	ssrR_putline(-100, 0, 100, 200, 0xffff0000);
+	
+//	printf("%u\n", (UINT_MAX  * (double)0.5F));
+
+}
diff --git a/src/example/02_cube.c b/src/example/02_cube.c
new file mode 100644
index 0000000..caedd64
--- /dev/null
+++ b/src/example/02_cube.c
@@ -0,0 +1,73 @@
+#include "example.h"
+
+/*正方体*/
+Vec3 verts[] = {
+	// front face 
+	{1, 1, 1}, {-1, 1, 1}, {-1, -1, 1}, {1, -1, 1},
+	// back face
+	{1, 1, -1}, {-1, 1, -1}, {-1, -1, -1}, {1, -1, -1},
+};
+
+Color colors[] = {
+	0xffff0000, 0xff00ff00, 0xffff00ff, 0xff00ffff,
+	0xff0000ff, 0xff000000, 0xffffff00, 0xffffffff,
+};
+
+int cube[] = {
+	0, 2, 1, 0, 3, 2, 
+	1, 2, 5, 2, 6, 5, 
+	4, 5, 6, 4, 6, 7, 
+	0, 4, 7, 0, 7, 3, 
+	0, 1, 4, 1, 5, 4, 
+	2, 3, 6, 3, 7, 6
+};
+Mat4 m;
+void onloadcube(void* data) {
+}
+
+void oneventcube(void* data) {
+	SDL_Event* e = (SDL_Event*)data;
+}
+
+float _t = 0;
+
+void onupdatecube(void*data) {
+	uint dt = *(uint*)data;
+	ssr_matrixmode(MATRIX_MODEL);
+	ssr_loadidentity();
+	ssr_translate(0, 0, -3);
+	ssr_rotate(360 * sin(_t += 0.001f), 1, 1, 1);
+	ssr_matrixmode(MATRIX_PROJECTION);
+	ssr_loadidentity();
+	ssr_perspective(100 + 20 * sin(_t * 10), 1.25f, -0.1f, -100);
+	ssr_matrixmode(MATRIX_VIEW);
+	ssr_loadidentity();
+	Vec3 pos = { 0,0,0 }, target = { 0,0,-1 }, up = { 0,1,0 };
+	ssr_lookat(&pos, &target, &up);
+	ssr_getmvp(&m);
+}
+
+void ondrawcube(void*data) {
+	ssr_clearcolor(0);
+	Vec2 proj[8];
+
+	for (int i = 0; i < 8; ++i) {
+		Vec4 v = { verts[i].x, verts[i].y ,verts[i].z ,1 }, temp;
+		mat4_applytovec4(&m, &v, &temp);
+		temp.x /= temp.w;
+		temp.y /= temp.w;
+		temp.z /= temp.w;
+		//vec4_print(&temp);
+		proj[i].x = temp.x;
+		proj[i].y = temp.y;
+	}
+	for (int j = 1; j < sizeof(cube) / sizeof(int); ++j) {
+		int fromx = proj[cube[j]].x * 250.f + 250, fromy = 200 - proj[cube[j]].y * 200.f;
+		int tox = proj[cube[j - 1]].x * 250.f + 250, toy = 200 - proj[cube[j - 1]].y * 200.f;
+		ssrR_putline(fromx, fromy , tox, toy, 0xffff0000);
+	}
+
+	Vec2 v1 = { 0, 0 }, v2 = { 3, 1 }, v3 = {1, 5};
+	float area = ssrR_area(&v1, &v2, &v3);
+	printf("%f\n", area);
+}
diff --git a/src/example/03_texture.c b/src/example/03_texture.c
new file mode 100644
index 0000000..fb5f0f2
--- /dev/null
+++ b/src/example/03_texture.c
@@ -0,0 +1,60 @@
+#include "example.h"
+
+static int cube[] = {
+	0, 1, 2, 0, 2, 3,
+	1, 5, 2, 2, 5, 6,
+	4, 6, 5, 4, 7, 6,
+	0, 3, 7, 0, 7, 4,
+	0, 4, 1, 1, 4, 5,
+	2, 6, 3, 3, 6, 7
+};
+
+static Vert verts[] = {
+	{0, {1, 1, 1}, zerovec3, zerovec3, zerovec2, 0xffff0000},
+	{1, {-1, 1, 1}, zerovec3, zerovec3, zerovec2,0xff00ff00},
+	{2, {-1, -1, 1}, zerovec3, zerovec3, zerovec2, 0xff0000ff},
+	{3, {1, -1, 1}, zerovec3, zerovec3, zerovec2, 0xffff00ff},
+	{4, {1, 1, -1}, zerovec3, zerovec3, zerovec2, 0xffaa28aa},
+	{5, {-1, 1, -1}, zerovec3, zerovec3, zerovec2,0xffFFC58E},
+	{6, {-1, -1, -1}, zerovec3, zerovec3, zerovec2, 0xffA100FF},
+	{7, {1, -1, -1}, zerovec3, zerovec3, zerovec2, 0xffFAFF00},
+};
+
+extern Program ssr_built_in_shader_unlit;
+
+void onloadtexture(void* data) {
+	ssr_matrixmode(MATRIX_PROJECTION);
+	ssr_loadidentity();
+	ssr_perspective(90, 1.25f, 0.1f, 10);
+	//ssr_ortho(-5, 5, -4, 4, 0.1, 10);
+	ssr_matrixmode(MATRIX_VIEW);
+	ssr_loadidentity();
+	Vec3 p = { 0, 0, 0 }, target = { 0, 0, -1 };
+	ssr_lookat(&p, &target, &vec3up);
+
+	ssr_bindvertices(verts, 8, cube, 12);
+	ssr_useprogram(&ssr_built_in_shader_unlit);
+	ssr_enable(ENABLEMASK_BACKFACECULL);
+	ssr_enable(ENABLEMASK_DEPTHTEST);
+}
+
+void oneventtexture(void* data) {
+	SDL_Event* e = (SDL_Event*)data;
+}
+
+static float _t = 0;
+
+void onupdatetexture(void*data) {
+	uint dt = *(uint*)data;
+	ssr_matrixmode(MATRIX_MODEL);
+	ssr_loadidentity();
+	ssr_translate(0, 0, -2);
+	ssr_rotate(_t -= dt / 50.f, 1, 1, 1);
+}
+
+void ondrawtexture(void*data) {
+	ssr_clearcolor(0x00);
+	ssr_cleardepth();
+
+	ssr_draw(PRIMITIVE_TRIANGLE);
+}
diff --git a/src/example/04_bloom.c b/src/example/04_bloom.c
new file mode 100644
index 0000000..e69de29
--- /dev/null
+++ b/src/example/04_bloom.c
diff --git a/src/example/05_fog.c b/src/example/05_fog.c
new file mode 100644
index 0000000..e69de29
--- /dev/null
+++ b/src/example/05_fog.c
diff --git a/src/extend/model.c b/src/extend/model.c
new file mode 100644
index 0000000..e69de29
--- /dev/null
+++ b/src/extend/model.c
diff --git a/src/extend/model.h b/src/extend/model.h
new file mode 100644
index 0000000..e69de29
--- /dev/null
+++ b/src/extend/model.h
diff --git a/src/extend/skeleton.c b/src/extend/skeleton.c
new file mode 100644
index 0000000..e69de29
--- /dev/null
+++ b/src/extend/skeleton.c
diff --git a/src/extend/skeleton.h b/src/extend/skeleton.h
new file mode 100644
index 0000000..e061506
--- /dev/null
+++ b/src/extend/skeleton.h
@@ -0,0 +1,6 @@
+#ifndef _SOFTSHADEROOM_SKELETON_H_
+#define _SOFTSHADEROOM_SKELETON_H_
+
+
+
+#endif
\ No newline at end of file
diff --git a/src/math/mat.c b/src/math/mat.c
new file mode 100644
index 0000000..84a5a41
--- /dev/null
+++ b/src/math/mat.c
@@ -0,0 +1,755 @@
+#include <math.h>
+#include <stdio.h>
+#include <string.h>
+
+#include "math.h"
+#include "../util/assert.h"
+#include "../core/mem.h"
+
+
+static Mat4 sharedMat;
+static Mat4 sharedMat2;
+static Vec4 sharedVec4;
+
+Mat4 mat4identity = {
+	1,0,0,0,
+	0,1,0,0,
+	0,0,1,0,
+	0,0,0,1
+};
+
+#define shrmat(p) \
+do{\
+sharedMat = *p;\
+p = &sharedMat;\
+}while(0)
+
+#define shrmat2(p) \
+do{\
+sharedMat2 = *p;\
+p = &sharedMat2;\
+}while(0)
+
+void mat4_tostring(Mat4* m, char str[]) {
+	ssrM_zero(str, sizeof(str));
+	for (int r = 0; r < 4; ++r) {
+		for (int c = 0; c < 4; ++c) {
+			sprintf(str, "%8.3f ", MAT(m, r, c) == -0 ? +0 : MAT(m, r, c));
+			str += strlen(str);
+		}
+		if(r != 3) sprintf(str, "\n");
+		str += strlen(str);
+	}
+}
+
+void mat4_print(Mat4* m) {
+	mat4_tostring(m, printbuffer);
+	printf("\n%s\n", printbuffer);
+}
+
+void mat4_zero(Mat4* out) {
+	ssr_assert(out);
+	ssrM_zero(out, sizeof(Mat4));
+}
+
+void mat4_setidentity(Mat4* out) {
+	ssr_assert(out);
+	mat4_zero(out);
+	out->e00 = 1; 
+	out->e11 = 1; 
+	out->e22 = 1; 
+	out->e33 = 1;
+}
+
+void mat4_setortho(float l, float r, float b, float t, float n, float f, Mat4* out) {
+	ssr_assert(out);
+	mat4_zero(out);
+	out->e00 = 2 / (r - l);
+	out->e03 = -(r + l) / (r - l);
+	out->e11 = 2 / (t - b);
+	out->e13 = -(t + b) / (t - b);
+	out->e22 = -2 / (f - n);
+	out->e23 = -(f + n) / (f - n);
+	out->e33 = 1;
+}
+
+void mat4_setfrustum(float l, float r, float b, float t, float n, float f, Mat4* out) {
+	ssr_assert(out);
+	mat4_zero(out);
+	out->e00 = (2.f * n) / (r - l);
+	out->e02 = (r + l) / (r - l);
+	out->e11 = 2.f * n / (t - b);
+	out->e12 = (t + b) / (t - b);
+	out->e22 = -(f + n) / (f - n);
+	out->e23 = -2.f * f * n / (f - n);
+	out->e32 = -1;
+}
+
+void mat4_setperspective(float _fov, float aspect, float near, float far, Mat4* out) {
+	float fov = _fov * PI / 180.f;
+	float tanf = tan(fov * 0.5);
+	ssr_assert(fov > 0 && aspect > 0);
+	ssr_assert(near > 0 && far > 0 && near != far);
+	mat4_setfrustum(
+	  -near*tanf*aspect,
+		near*tanf*aspect,
+		-near*tanf,
+		near*tanf, 
+		near, 
+		far,
+		out
+	);
+}
+
+static float _mul(float* r, float* c) {
+	return c[0] * r[0] + c[1] * r[4] + c[2] * r[8] + c[3] * r[12];
+}
+
+#define mul(r, c) _mul(&MAT(m1,r,0), &MAT(m2,0,c))
+
+void mat4_multiply(Mat4* m1, Mat4* m2, Mat4* out) {
+	ssr_assert(m1 && m2 && out);
+	if (mat4_isidentity(m1)) { if(m2 != out) *out = *m2; return; }
+	if (mat4_isidentity(m2)) { if(m1 != out) *out = *m1; return; }
+	if (m1 == out) shrmat(m1);
+	if (m2 == out) shrmat2(m2);
+
+	out->e00 = mul(0, 0); out->e01 = mul(0, 1); out->e02 = mul(0, 2); out->e03 = mul(0, 3);
+	out->e10 = mul(1, 0); out->e11 = mul(1, 1); out->e12 = mul(1, 2); out->e13 = mul(1, 3);
+	out->e20 = mul(2, 0); out->e21 = mul(2, 1); out->e22 = mul(2, 2); out->e23 = mul(2, 3);
+	out->e30 = mul(3, 0); out->e31 = mul(3, 1); out->e32 = mul(3, 2); out->e33 = mul(3, 3);
+}
+
+void mat4_setscale(float kx, float ky, float kz, Mat4* out) {
+	ssr_assert(out);
+	mat4_zero(out);
+	out->e00 = kx;
+	out->e11 = ky;
+	out->e22 = kz;
+	out->e33 = 1;
+}
+
+void mat4_setposition(float x, float y, float z, Mat4* out) {
+	ssr_assert(out);
+	mat4_setidentity(out);
+	out->e03 = x;
+	out->e13 = y;
+	out->e23 = z;
+}
+
+void mat4_setrotatez(float angle, Mat4* out) {
+	ssr_assert(out);
+	mat4_setidentity(out);
+	angle = radians(angle);
+	float s = sin(angle), c = cos(angle);
+	out->e00 = c; out->e01 = -s;
+	out->e10 = s; out->e11 = c;
+}
+
+void mat4_setrotatex(float angle, Mat4* out) {
+	ssr_assert(out);
+	mat4_setidentity(out);
+	angle = radians(angle);
+	float s = sin(angle), c = cos(angle);
+	out->e11 = c; out->e12 = -s; 
+	out->e21 = s; out->e22 = c;
+}
+
+void mat4_setrotatey(float angle, Mat4* out) {
+	ssr_assert(out);
+	mat4_setidentity(out);
+	angle = radians(angle);
+	float s = sin(angle), c = cos(angle);
+	out->e00 = c; out->e02 = s;
+	out->e20 = -s; out->e22 = c;
+}
+
+/*https://www.geometrictools.com/Documentation/EulerAngles.pdf*/
+void mat4_setrotate(float angleX, float angleY, float angleZ, Mat4* out) {
+	ssr_assert(out);
+	mat4_setidentity(out);
+	angleX = radians(angleX); angleY = radians(angleY); angleZ = radians(angleZ);
+	float sx = sin(angleX), cx = cos(angleX);
+	float sy = sin(angleY), cy = cos(angleY);
+	float sz = sin(angleZ), cz = cos(angleZ);
+	out->e00 = cy * cz + sx * sy * sz;  out->e01 = cz * sx*sy - cy * sz; out->e02 = cx * sy;
+	out->e10 = cx * sz;                 out->e11 = cx * cz;              out->e12 = -sx;
+	out->e20 = -cz * sy + cy * sx * sz; out->e21 = cy * cz*sx + sy * sz; out->e22 = cx * cy;
+}
+
+void mat4_setaxisangle(Vec3* ax, float angle, Mat4* out) {
+	ssr_assert(ax && out);
+
+	float a = radians(angle);
+	float c = cos(a);
+	float s = sin(a);
+
+	Vec3 axis = *ax;
+	Vec3 temp;
+	vec3_normalize(&axis, &axis);
+	vec3_scale(&axis, 1 - c, &temp);
+
+	/*
+	rotation matrix 推导过程 https://zhuanlan.zhihu.com/p/56587491
+	X^2(1-c)+c, XY(1-c)-Zs, XZ(1-c)+Ys, 0
+	XY(1-c)+Zs, Y^2(1-c)+c, YZ(1-c)-Xs, 0
+	XZ(1-c)-Ys, YZ(1-c)+Xs, Z^2(1-c)+c, 0
+	0,          0,          0, 1
+	*/
+
+	mat4_setidentity(out);
+	out->m[0][0] = c + temp.x * axis.x;
+	out->m[0][1] = 0 + temp.x * axis.y + s * axis.z;
+	out->m[0][2] = 0 + temp.x * axis.z - s * axis.y;
+
+	out->m[1][0] = 0 + temp.y * axis.x - s * axis.z;
+	out->m[1][1] = c + temp.y * axis.y;
+	out->m[1][2] = 0 + temp.y * axis.z + s * axis.x;
+
+	out->m[2][0] = 0 + temp.z * axis.x + s * axis.y;
+	out->m[2][1] = 0 + temp.z * axis.y - s * axis.x;
+	out->m[2][2] = c + temp.z * axis.z;
+
+}
+
+void mat4_setorthonormalbias(Vec3* x, Vec3* y, Vec3* z, Mat4* out) {
+	ssr_assert(x && y && z);
+	mat4_setidentity(out);
+	Vec4 asix = { x->x, x->y, x->z, 0 };
+	Vec4 asiy = { y->x, y->y, y->z, 0 };
+	Vec4 asiz = { z->x, z->y, z->z, 0 };
+	out->colums[0] = asix;
+	out->colums[1] = asiy;
+	out->colums[2] = asiz;
+}
+
+bool mat4_isidentity(Mat4* m) {
+	ssr_assert(m);
+	//return memcmp(m, &mat4identity, sizeof(Mat4)) == 0;
+	return
+		compare(m->axisx.x, 1) && compare(m->axisx.y, 0) && compare(m->axisx.z,0) && compare(m->axisx.w, 0) &&
+		compare(m->axisy.x, 0) && compare(m->axisy.y, 1) && compare(m->axisy.z,0) && compare(m->axisy.w, 0) &&
+		compare(m->axisz.x, 0) && compare(m->axisz.y, 0) && compare(m->axisz.z,1) && compare(m->axisz.w, 0) &&
+		compare(m->pos.x, 0  ) && compare(m->pos.y, 0  ) && compare(m->pos.z, 0  ) &&compare( m->pos.w, 1);
+}
+
+bool mat4_isorthogonal(Mat4* m) {
+	ssr_assert(m);
+	Mat4 trans = {0}, res = { 0 };
+	mat4_transpose(m, &trans);
+	mat4_multiply(m, &trans, &res);
+	return mat4_isidentity(&res);
+}
+
+/*
+** 以z轴为准进行正交化，分为施密特正交化和叉乘正交化，施密特过程更加普遍，叉乘适用于三维空间，两种方法实际上等价
+** 如果用叉乘的方法，只需要关注yz，x通过叉乘得到
+*/
+void mat4_orthogonalize(Mat4* in, Mat4* out) {
+	ssr_assert(in && out);
+	if (in == out) {
+		shrmat(in);
+	}
+	
+	mat4_setidentity(out);
+	Vec4 z = in->basis.z;
+	vec3_normalize(&z, &z);
+	Vec4 y = in->basis.y;
+	Vec4 x = {0};
+	vec3_cross(&y, &z, &x);
+	vec3_normalize(&x, &x);
+	vec3_cross(&z, &x, &y);
+	out->basis.x = x;
+	out->basis.y = y;
+	out->basis.z = z;
+	
+	/*
+	mat4_setidentity(out);
+
+	Vec4 x = in->basis.x;
+	Vec4 y = in->basis.y;
+	Vec4 z = in->basis.z;
+	Vec3 temp, temp2;
+
+	vec3_normalize(&z, &z);
+	out->basis.z = z;
+
+	float dot = vec3_dot(&y, &z);
+	vec3_scale(&z, dot, &temp);
+	vec3_minus(&y, &temp, &y);
+	vec3_normalize(&y, &y);
+	out->basis.y = y;
+
+	vec3_cross(&y, &z, &out->basis.x);
+	*/
+	/*针对右手系调整basis.x的方向*/ 
+	/*https://math.stackexchange.com/questions/1847465/why-to-use-gram-schmidt-process-to-orthonormalise-a-basis-instead-of-cross-produ*/
+	/*由于需要针对右手系，这里不这样计算，因为可能要对结果进行翻转
+	dot = vec3_dot(&x, &z);
+	vec3_scale(&z, dot, &temp);
+	vec3_minus(&x, &temp, &temp2);
+	dot = vec3_dot(&x, &y);
+	vec3_scale(&y, dot, &temp);
+	vec3_minus(&temp2, &temp, &x);
+	vec3_normalize(&x, &x);
+	out->basis.x = x;
+	*/
+}
+
+bool mat4_setlookrotation(Vec3* view, Vec3* up, Mat4* out) {
+	ssr_assert(view && up && out);
+
+	/*正交化*/
+	float mag = vec3_magnitude(view);
+	if (mag < EPSILON) return 0;
+	Vec3 z;
+	vec3_scale(view, 1.f / mag, &z);
+
+	Vec3 x;
+	vec3_cross(up, &z, &x);
+	mag = vec3_magnitude(&x);
+	if (mag < EPSILON) return 0;
+	vec3_scale(&x, 1.f / mag, &x);
+
+	Vec3 y;
+	vec3_cross(&z, &x, &y);
+	mag = vec3_magnitude(&y);
+	if (!compare(mag, 1)) return 0;
+
+	mat4_setorthonormalbias(&x, &y, &z, out); /*xyz正交*/
+
+	return 1;
+}
+
+void mat4_applytovec4(Mat4* mat, Vec4* v, Vec4* out) {
+	ssr_assert(mat && v && out);
+	if (v == out) {
+		sharedVec4 = *v;
+		v = &sharedVec4;
+	}
+	out->x = mat->e00 * v->x + mat->e01 * v->y + mat->e02 * v->z + mat->e03 * v->w;
+	out->y = mat->e10 * v->x + mat->e11 * v->y + mat->e12 * v->z + mat->e13 * v->w;
+	out->z = mat->e20 * v->x + mat->e21 * v->y + mat->e22 * v->z + mat->e23 * v->w;
+	out->w = mat->e30 * v->x + mat->e31 * v->y + mat->e32 * v->z + mat->e33 * v->w;
+}
+
+#define trans(r, c) out->e##r##c = m->e##c##r
+
+void mat4_transpose(Mat4* m, Mat4* out) {
+	ssr_assert(m && out);
+	if (m == out) shrmat(m);
+
+	trans(0, 0); trans(0, 1); trans(0, 2); trans(0, 3);
+	trans(1, 0); trans(1, 1); trans(1, 2); trans(1, 3);
+	trans(2, 0); trans(2, 1); trans(2, 2); trans(2, 3);
+	trans(3, 0); trans(3, 1); trans(3, 2); trans(3, 3);
+}
+
+/*
+** 使用高斯消元法计算任意矩阵的逆矩阵。针对不含投影的3D变换矩阵，应该使用
+** mat4_invertgeneral3d() 
+** 更快一些
+*/
+bool mat4_invertfull(Mat4* m, Mat4* out) {
+	ssr_assert(m && out);
+
+#define _m(r, c) MAT(m, r, c)
+	float wtmp[4][8] = {
+		{ /*M*/ _m(0,0), _m(0, 1), _m(0, 2), _m(0, 3), /*I*/ 1, 0, 0, 0 },
+		{ /*M*/ _m(1,0), _m(1, 1), _m(1, 2), _m(1, 3), /*I*/ 0, 1, 0, 0 },
+		{ /*M*/ _m(2,0), _m(2, 1), _m(2, 2), _m(2, 3), /*I*/ 0, 0, 1, 0 },
+		{ /*M*/ _m(3,0), _m(3, 1), _m(3, 2), _m(3, 3), /*I*/ 0, 0, 0, 1 },
+	};
+#undef _m
+	float m0, m1, m2, m3, s;
+	float *r0, *r1, *r2, *r3;
+	r0 = wtmp[0], r1 = wtmp[1], r2 = wtmp[2], r3 = wtmp[3];
+#define SWAP_ROWS(a, b) { float *_tmp = a; (a)=(b); (b)=_tmp; }
+
+	//#define optimize(block) if(s!=0.f){block}
+#define optimize(block) block
+
+	/* choose pivot - or die */
+	if (absf(r3[0]) > absf(r2[0])) SWAP_ROWS(r3, r2);
+	if (absf(r2[0]) > absf(r1[0])) SWAP_ROWS(r2, r1);
+	if (absf(r1[0]) > absf(r0[0])) SWAP_ROWS(r1, r0);
+	if (0.0f == r0[0]) return 0;
+
+	/* eliminate first variable */
+	m1 = r1[0] / r0[0]; m2 = r2[0] / r0[0]; m3 = r3[0] / r0[0];
+	s = r0[1]; optimize(r1[1] -= m1 * s; r2[1] -= m2 * s; r3[1] -= m3 * s; )
+	s = r0[2]; optimize(r1[2] -= m1 * s; r2[2] -= m2 * s; r3[2] -= m3 * s; )
+	s = r0[3]; optimize(r1[3] -= m1 * s; r2[3] -= m2 * s; r3[3] -= m3 * s; )
+	s = r0[4]; optimize(r1[4] -= m1 * s; r2[4] -= m2 * s; r3[4] -= m3 * s; )
+	s = r0[5]; optimize(r1[5] -= m1 * s; r2[5] -= m2 * s; r3[5] -= m3 * s; )
+	s = r0[6]; optimize(r1[6] -= m1 * s; r2[6] -= m2 * s; r3[6] -= m3 * s; )
+	s = r0[7]; optimize(r1[7] -= m1 * s; r2[7] -= m2 * s; r3[7] -= m3 * s; )
+
+	/* choose pivot - or die */
+	if (absf(r3[1]) > absf(r2[1])) SWAP_ROWS(r3, r2);
+	if (absf(r2[1]) > absf(r1[1])) SWAP_ROWS(r2, r1);
+	if (0.0F == r1[1]) return 0;
+
+	/* eliminate second variable */
+	m2 = r2[1] / r1[1]; m3 = r3[1] / r1[1];
+	s = r1[2]; optimize(r2[2] -= m2 * s; r3[2] -= m3 * s; )
+	s = r1[3]; optimize(r2[3] -= m2 * s; r3[3] -= m3 * s; )
+	s = r1[4]; optimize(r2[4] -= m2 * s; r3[4] -= m3 * s; )
+	s = r1[5]; optimize(r2[5] -= m2 * s; r3[5] -= m3 * s; )
+	s = r1[6]; optimize(r2[6] -= m2 * s; r3[6] -= m3 * s; )
+	s = r1[7]; optimize(r2[7] -= m2 * s; r3[7] -= m3 * s; )
+
+	/* choose pivot - or die */
+	if (absf(r3[2])>absf(r2[2])) SWAP_ROWS(r3, r2);
+	if (0.0F == r2[2]) return 0;
+
+	/* eliminate third variable */
+	m3 = r3[2] / r2[2];
+	s = r2[3]; optimize(r3[3] -= m3 * s; )
+	s = r2[4]; optimize(r3[4] -= m3 * s; )
+	s = r2[5]; optimize(r3[5] -= m3 * s; )
+	s = r2[6]; optimize(r3[6] -= m3 * s; )
+	s = r2[7]; optimize(r3[7] -= m3 * s; )
+
+#undef optimize
+
+	/* last check */
+	if (0.0F == r3[3]) return 0;
+
+	s = 1.0F / r3[3];             /* now back substitute row 3 */
+	r3[4] *= s; r3[5] *= s; r3[6] *= s; r3[7] *= s;
+
+	m2 = r2[3];                 /* now back substitute row 2 */
+	s = 1.0F / r2[2];
+	r2[4] = s * (r2[4] - r3[4] * m2), r2[5] = s * (r2[5] - r3[5] * m2),
+	r2[6] = s * (r2[6] - r3[6] * m2), r2[7] = s * (r2[7] - r3[7] * m2);
+	m1 = r1[3];
+	r1[4] -= r3[4] * m1, r1[5] -= r3[5] * m1,
+	r1[6] -= r3[6] * m1, r1[7] -= r3[7] * m1;
+	m0 = r0[3];
+	r0[4] -= r3[4] * m0, r0[5] -= r3[5] * m0,
+	r0[6] -= r3[6] * m0, r0[7] -= r3[7] * m0;
+
+	m1 = r1[2];                 /* now back substitute row 1 */
+	s = 1.0F / r1[1];
+	r1[4] = s * (r1[4] - r2[4] * m1), r1[5] = s * (r1[5] - r2[5] * m1),
+	r1[6] = s * (r1[6] - r2[6] * m1), r1[7] = s * (r1[7] - r2[7] * m1);
+	m0 = r0[2];
+	r0[4] -= r2[4] * m0, r0[5] -= r2[5] * m0,
+	r0[6] -= r2[6] * m0, r0[7] -= r2[7] * m0;
+
+	m0 = r0[1];                 /* now back substitute row 0 */
+	s = 1.0F / r0[0];
+	r0[4] = s * (r0[4] - r1[4] * m0), r0[5] = s * (r0[5] - r1[5] * m0),
+	r0[6] = s * (r0[6] - r1[6] * m0), r0[7] = s * (r0[7] - r1[7] * m0);
+
+	out->e00 = r0[4]; out->e01 = r0[5]; out->e02 = r0[6]; out->e03 = r0[7];
+	out->e10 = r1[4]; out->e11 = r1[5]; out->e12 = r1[6]; out->e13 = r1[7];
+	out->e20 = r2[4]; out->e21 = r2[5]; out->e22 = r2[6]; out->e23 = r2[7];
+	out->e30 = r3[4]; out->e31 = r3[5]; out->e32 = r3[6]; out->e33 = r3[7];
+
+#undef SWAP_ROWS
+
+	return 1;
+}
+
+/*
+** 对只包含基本3D变换的矩阵进行变换，先计算左上角3x3的RS矩阵的逆（通过伴随矩阵），然后
+** 乘上平移矩阵的逆矩阵，即
+** M^-1 = (T(RS))^-1 = (RS)^-1 * T^-1
+*/
+bool mat4_invertgeneral3d(Mat4* in, Mat4* out) {
+	ssr_assert(in && out);
+	if (in == out) shrmat(in);
+
+	mat4_setidentity(out);
+
+	/*计算左上角3x3矩阵的行列式*/
+	float pos = 0, neg = 0, t;
+	float det;
+
+	t = in->e00 * in->e11 * in->e22;
+	if (t >= 0) pos += t; else neg += t;
+	t = in->e10 * in->e21 * in->e02;
+	if (t >= 0) pos += t; else neg += t;
+	t = in->e20 * in->e01 * in->e12;
+	if (t >= 0) pos += t; else neg += t;
+
+	t = -in->e20 * in->e11 * in->e02;
+	if (t >= 0) pos += t; else neg += t;
+	t = -in->e10 * in->e01 * in->e22;
+	if (t >= 0) pos += t; else neg += t;
+	t = -in->e00 * in->e21 * in->e12;
+	if (t >= 0) pos += t; else neg += t;
+
+	det = pos + neg;
+
+	if (det * det < 1e-25)
+		return 0; /*行列式为0*/
+
+	det = 1.f / det; 
+	MAT(out, 0, 0) = ((MAT(in, 1, 1)*MAT(in, 2, 2) - MAT(in, 2, 1)*MAT(in, 1, 2))*det);
+	MAT(out, 0, 1) = (-(MAT(in, 0, 1)*MAT(in, 2, 2) - MAT(in, 2, 1)*MAT(in, 0, 2))*det);
+	MAT(out, 0, 2) = ((MAT(in, 0, 1)*MAT(in, 1, 2) - MAT(in, 1, 1)*MAT(in, 0, 2))*det);
+	MAT(out, 1, 0) = (-(MAT(in, 1, 0)*MAT(in, 2, 2) - MAT(in, 2, 0)*MAT(in, 1, 2))*det);
+	MAT(out, 1, 1) = ((MAT(in, 0, 0)*MAT(in, 2, 2) - MAT(in, 2, 0)*MAT(in, 0, 2))*det);
+	MAT(out, 1, 2) = (-(MAT(in, 0, 0)*MAT(in, 1, 2) - MAT(in, 1, 0)*MAT(in, 0, 2))*det);
+	MAT(out, 2, 0) = ((MAT(in, 1, 0)*MAT(in, 2, 1) - MAT(in, 2, 0)*MAT(in, 1, 1))*det);
+	MAT(out, 2, 1) = (-(MAT(in, 0, 0)*MAT(in, 2, 1) - MAT(in, 2, 0)*MAT(in, 0, 1))*det);
+	MAT(out, 2, 2) = ((MAT(in, 0, 0)*MAT(in, 1, 1) - MAT(in, 1, 0)*MAT(in, 0, 1))*det);
+
+	// 乘T^-1
+	MAT(out, 0, 3) = -(MAT(in, 0, 3) * MAT(out, 0, 0) +
+		MAT(in, 1, 3) * MAT(out, 0, 1) +
+		MAT(in, 2, 3) * MAT(out, 0, 2));
+	MAT(out, 1, 3) = -(MAT(in, 0, 3) * MAT(out, 1, 0) +
+		MAT(in, 1, 3) * MAT(out, 1, 1) +
+		MAT(in, 2, 3) * MAT(out, 1, 2));
+	MAT(out, 2, 3) = -(MAT(in, 0, 3) * MAT(out, 2, 0) +
+		MAT(in, 1, 3) * MAT(out, 2, 1) +
+		MAT(in, 2, 3) * MAT(out, 2, 2));
+
+	return 1;
+}
+
+void mat4_invertpos(Mat4* in, Mat4* out) {
+
+}
+
+void mat4_invertscale(Mat4* in, Mat4* out) {
+
+}
+
+void mat4_invertrot(Mat4* in, Mat4* out) {
+	ssr_assert(in && out);
+	mat4_transpose(in, out);
+}
+
+void mat4_settr(Vec3* pos, Quat* rot, Mat4* out) {
+	ssr_assert(pos && rot && out);
+	mat4_zero(out);
+	quat_tomat4(rot, out);
+	out->e03 = pos->x;
+	out->e13 = pos->y;
+	out->e23 = pos->z;
+}
+
+void mat4_settrs(Vec3* pos, Quat* rot, Vec3* scale, Mat4* out) {
+	ssr_assert(pos && rot && scale && out);
+	mat4_zero(out);
+	quat_tomat4(rot, out); /*pos*rot*scale的顺序*/
+	out->e00 *= scale->x; out->e01 *= scale->y; out->e02 *= scale->z;
+	out->e10 *= scale->x; out->e11 *= scale->y; out->e12 *= scale->z;
+	out->e20 *= scale->x; out->e21 *= scale->y; out->e22 *= scale->z;
+	out->e03 = pos->x;
+	out->e13 = pos->y;
+	out->e23 = pos->z;
+}
+
+void mat4_settrinverse(Vec3* pos, Quat* rot, Mat4* out) {
+	ssr_assert(pos && rot && out);
+	mat4_zero(out);
+	quat_invert(rot, rot);
+	quat_tomat4(rot, out);
+	Vec3 reverse = { -pos->x, -pos->y, -pos->z};
+	mat4_translate(out, &reverse, out); /* (TR)^-1 = R^-1*T^-1所以这里是右乘*/
+}
+
+void mat4_scale(Mat4* m, Vec3* scale, Mat4* out) {
+	ssr_assert(m && scale && out);
+	if (out != m) {
+		*out = *m;
+	}
+	/*
+	scale matrix 
+	x, 0, 0, 0,
+	0, y, 0, 0,
+	0, 0, z, 0, 
+	0, 0, 0, 1
+	*/
+	out->e00 *= scale->x;
+	out->e10 *= scale->x;
+	out->e20 *= scale->x;
+	out->e30 *= scale->x;
+
+	out->e01 *= scale->y;
+	out->e11 *= scale->y;
+	out->e21 *= scale->y;
+	out->e31 *= scale->y;
+	
+	out->e02 *= scale->z;
+	out->e12 *= scale->z;
+	out->e22 *= scale->z;
+	out->e32 *= scale->z;
+}
+
+void mat4_translate(Mat4* m, Vec3* pos, Mat4* out) {
+	ssr_assert(m && pos && out);
+	if (out != m) {
+		*out = *m;
+	}
+	/*
+	translate matrix 
+	1, 0, 0, x,
+	0, 1, 0, y,
+	0, 0, 1, z,
+	0, 0, 0, 1,
+	*/
+	out->e03 = out->e00 * pos->x + out->e01 * pos->y + out->e02 * pos->z + out->e03;
+	out->e13 = out->e10 * pos->x + out->e11 * pos->y + out->e12 * pos->z + out->e13;
+	out->e23 = out->e20 * pos->x + out->e21 * pos->y + out->e22 * pos->z + out->e23;
+	out->e33 = out->e30 * pos->x + out->e31 * pos->y + out->e32 * pos->z + out->e33;
+}
+
+void mat4_rotate(Mat4* m, float angle, Vec3* ax, Mat4* out) {
+	ssr_assert(m && ax && out);
+	Mat4 rot; 
+	mat4_setaxisangle(ax, angle, &rot);
+	mat4_multiply(m, &rot, out);
+}
+
+void mat4_decomposetrs(Mat4* src, Vec3* pos, Quat* quat, Vec3* scale) {
+	ssr_assert(src && pos && quat && scale); 
+
+	Vec3* x = &src->colums[0];
+	Vec3* y = &src->colums[1];
+	Vec3* z = &src->colums[2];
+	Vec3* w = &src->colums[3];
+
+	*pos = *w;
+
+	quat_setlookrotation(z, y, quat);
+
+	scale->x = vec3_magnitude(x);
+	scale->y = vec3_magnitude(y);
+	scale->z = vec3_magnitude(z);
+}
+
+static void MakePositive(Euler* euler) {/*弧度制欧拉角*/
+	const float negativeFlip = -0.0001F;
+	const float positiveFlip = (PI * 2.0F) - 0.0001F;
+
+	if (euler->x < negativeFlip)
+		euler->x += 2.0 * PI;
+	else if (euler->x > positiveFlip)
+		euler->x -= 2.0 * PI;
+
+	if (euler->y < negativeFlip)
+		euler->y += 2.0 * PI;
+	else if (euler->y > positiveFlip)
+		euler->y -= 2.0 * PI;
+
+	if (euler->z < negativeFlip)
+		euler->z += 2.0 * PI;
+	else if (euler->z > positiveFlip)
+		euler->z -= 2.0 * PI;
+}
+
+static void SanitizeEuler(Euler* e) {/*弧度制欧拉角*/
+	MakePositive(e);
+}
+
+/*from unity src*/
+bool mat4_toeuler(Mat4* in, Euler* out) {
+	ssr_assert(in && out);
+	// from http://www.geometrictools.com/Documentation/EulerAngles.pdf
+	// YXZ order
+	if (MAT(in, 1, 2) < 0.999F) // some fudge for imprecision
+	{
+		if (MAT(in, 1, 2) > -0.999F) // some fudge for imprecision
+		{
+			out->x = asin(-MAT(in, 1, 2));
+			out->y = atan2(MAT(in, 0, 2), MAT(in, 2, 2));
+			out->z = atan2(MAT(in, 1, 0), MAT(in, 1, 1));
+			//euler_rad2deg(out, out);
+			SanitizeEuler(out);
+			euler_rad2deg(out, out);
+			return 1;
+		}
+		else
+		{
+			// WARNING.  Not unique.  YA - ZA = atan2(r01,r00)
+			out->x = PI * 0.5F;
+			out->y = atan2(MAT(in, 0, 1), MAT(in, 0, 0));
+			out->z = 0.0F;
+			//euler_rad2deg(out, out);
+			SanitizeEuler(out);
+			euler_rad2deg(out, out);
+			return 0;
+		}
+	}
+	else
+	{
+		// WARNING.  Not unique.  YA + ZA = atan2(-r01,r00)
+		out->x = -PI * 0.5F;
+		out->y = atan2(-MAT(in, 0, 1), MAT(in, 0, 0));
+		out->z = 0.0F;
+		//euler_rad2deg(out, out);
+		SanitizeEuler(out);
+		euler_rad2deg(out, out);
+		return 0;
+	}
+}
+
+/*from unity src*/
+/*https://www.euclideanspace.com/maths/geometry/rotations/conversions/matrixToQuaternion/*/
+void mat4_toquat(Mat4* in, Quat* out) {
+	// Algorithm in Ken Shoemake's article in 1987 SIGGRAPH course notes
+	// article "Quaternionf Calculus and Fast Animation".
+	float fTrace = MAT(in, 0, 0) + MAT(in, 1, 1) + MAT(in, 2, 2);
+	float fRoot;
+
+	if (fTrace > 0.0f)
+	{
+		// |w| > 1/2, may as well choose w > 1/2
+		fRoot = sqrt(fTrace + 1.0f);  // 2w
+		out->w = 0.5f*fRoot;
+		fRoot = 0.5f / fRoot;  // 1/(4w)
+		out->x = (MAT(in, 2, 1) - MAT(in, 1, 2))*fRoot;
+		out->y = (MAT(in, 0, 2) - MAT(in, 2, 0))*fRoot;
+		out->z = (MAT(in, 1, 0) - MAT(in, 0, 1))*fRoot;
+	}
+	else
+	{
+		// |w| <= 1/2
+		int s_iNext[3] = { 1, 2, 0 };
+		int i = 0;
+		if (MAT(in, 1, 1) > MAT(in, 0, 0))
+			i = 1;
+		if (MAT(in, 2, 2) > MAT(in, i, i))
+			i = 2;
+		int j = s_iNext[i];
+		int k = s_iNext[j];
+
+		fRoot = sqrt(MAT(in, i, i) - MAT(in, j, j) - MAT(in, k, k) + 1.0f);
+		float* apkQuat[3] = { &out->x, &out->y, &out->z };
+		ssr_assert(fRoot >= EPSILON);
+		*apkQuat[i] = 0.5f*fRoot;
+		fRoot = 0.5f / fRoot;
+		out->w = (MAT(in, k, j) - MAT(in, j, k)) * fRoot;
+		*apkQuat[j] = (MAT(in, j, i) + MAT(in, i, j))*fRoot;
+		*apkQuat[k] = (MAT(in, k, i) + MAT(in, i, k))*fRoot;
+	}
+	quat_normalize(out, out);
+}
+
+void mat3_applytovec3(Mat3* m, Vec3* v, Vec3* out) {
+	ssr_assert(m && v && out);
+	out->x = m->e00 * v->x + m->e01 * v->y + m->e02 * v->z;
+	out->y = m->e10 * v->x + m->e11 * v->y + m->e12 * v->z;
+	out->z = m->e20 * v->x + m->e21 * v->y + m->e22 * v->z;
+}
+
+void mat23_applytovec3(Mat23* m, Vec3* v, Vec2* out) {
+	ssr_assert(m && v && out);
+	out->x = m->e00 * v->x + m->e01 * v->y + m->e02 * v->z;
+	out->y = m->e10 * v->x + m->e11 * v->y + m->e12 * v->z;
+}
+
+void mat43_applytovec3(Mat43* m, Vec3* v, Vec4* out) {
+	ssr_assert(m && v && out);
+	out->x = m->e00 * v->x + m->e01 * v->y + m->e02 * v->z;
+	out->y = m->e10 * v->x + m->e11 * v->y + m->e12 * v->z;
+	out->z = m->e20 * v->x + m->e21 * v->y + m->e22 * v->z;
+	out->w = m->e30 * v->x + m->e31 * v->y + m->e32 * v->z;
+}
\ No newline at end of file
diff --git a/src/shaders/common_header.h b/src/shaders/common_header.h
new file mode 100644
index 0000000..848fe98
--- /dev/null
+++ b/src/shaders/common_header.h
@@ -0,0 +1,7 @@
+#ifndef _SOFTSHADEROOM_COMMON_HEADER_H_
+#define _SOFTSHADEROOM_COMMON_HEADER_H_
+
+#include "../core/shader.h"
+
+
+#endif
\ No newline at end of file
diff --git a/src/shaders/default.c b/src/shaders/default.c
new file mode 100644
index 0000000..b1345e0
--- /dev/null
+++ b/src/shaders/default.c
@@ -0,0 +1,20 @@
+#include "common_header.h"
+
+/*uniforms*/
+#define lightdir UV3(0)
+#define shinees  UN(0)
+
+/*vertex output*/
+#define uv reg_v2_00
+
+static void vert(UniformCollection* uniforms, VertexShaderIn* in, Vec4* homocoord) {
+	static Vec4 p; p.xyz = in->vertex->position; p.w = 1;
+	mat4_applytovec4(uniforms->mvp, &p, homocoord);
+}
+
+static bool frag(UniformCollection* uniforms, FragmentShaderIn* in, Color* color) {
+
+	return 1;
+}
+
+Program ssr_built_in_shader_default = { vert, frag };
\ No newline at end of file
diff --git a/src/shaders/unlit.c b/src/shaders/unlit.c
new file mode 100644
index 0000000..ec6949f
--- /dev/null
+++ b/src/shaders/unlit.c
@@ -0,0 +1,23 @@
+#include "common_header.h"
+
+#define model_matrix UM4(0)
+
+#define world_normal reg_v3_00
+#define vcolor  reg_v4_00
+
+static void vert(UniformCollection* uniforms, VertexShaderIn* in, Vec4* clipcoord) {
+	static Vec4 p; p.xyz = in->vertex->position; p.w = 1;
+	mat4_applytovec4(uniforms->mvp, &p, clipcoord);
+	model_matrix;
+}
+
+static bool frag(UniformCollection* uniforms, FragmentShaderIn* in, Color* color) {
+	*color = in->color;
+	world_normal;
+	return 1;
+}
+
+Program ssr_built_in_shader_unlit = { 
+	vert, frag,
+	VARYING_COLOR
+};