Add rendering functions (crashes)

example/source/main.cpp

@@ -46,10 +46,17 @@ class example : public amy::app {
  public:
   example() {
     amy::ctru::init();
+    consoleInit(GFX_BOTTOM, NULL);
     amy::c3d::init();
     m_top = amy::c3d::createScreen(GFX_TOP, GFX_LEFT);
     amy::iron::init();
-  };
+    dl = new amy::iron::drawlist();
+    dl->drawSolid();
+    // throw std::runtime_error(std::format(
+    //     "solid tex: {:#08x}\nsize: {}\nptr: {:#08x}",
+    //     (amy::ui)amy::iron::whiteTex(), amy::iron::whiteTex()->size(),
+    //     (amy::ui)amy::iron::whiteTex()->ptr()));
+  }; 
   ~example() {
     amy::c3d::deleteScreen(m_top);
     amy::c3d::deinit();
@@ -59,13 +66,17 @@ class example : public amy::app {
     amy::c3d::startFrame();
     m_top->startDraw();
     m_top->clear();
+    dl->drawRectFilled(0, 50, 0xff00ff00);
     amy::iron::newFrame();
     amy::iron::drawOn(m_top);
+    amy::iron::draw(dl->data());
+    dl->clear();
     amy::c3d::endFrame();
   }
 
  private:
   amy::c3d::screen* m_top = nullptr;
+  amy::iron::drawlist* dl = nullptr;
 };
 
 int main() {

include/amethyst/c3d.hpp

@@ -83,7 +83,8 @@ class c3d {
   static void deleteScreen(screen* screen);
   static void drawArrays(int start, int count,
                          GPU_Primitive_t prim = GPU_TRIANGLES);
-  static void drawElements(int count, const void* idx_ptr, int type = GPU_SHORT,
+  static void drawElements(int count, const void* idx_ptr,
+                           int type = C3D_UNSIGNED_SHORT,
                            GPU_Primitive_t prim = GPU_TRIANGLES);
   static void depthTest(bool on, GPU_TESTFUNC func = GPU_GREATER,
                         GPU_WRITEMASK mask = GPU_WRITE_ALL);

include/amethyst/iron.hpp

@@ -1,10 +1,10 @@
 #pragma once
 
-#include <algorithm>
 #include <amethyst/c3d.hpp>
 #include <amethyst/linearAlloc.hpp>
 #include <amethyst/maths/mat.hpp>
 #include <amethyst/maths/vec.hpp>
+#include <amethyst/rect.hpp>
 #include <amethyst/texture.hpp>
 #include <amethyst/types.hpp>
 
@@ -12,18 +12,23 @@ namespace amy {
 class iron {
  public:
   struct vertex {
-    vertex(float x, float y, float u, float v, u32 clr) {
+    vertex(float x, float y, float u, float v, ui clr) {
       pos.x = x;
       pos.y = y;
       uv.x = x;
       uv.y = y;
       color = clr;
     }
+    vertex(const fvec2& pos, const fvec2& uv, ui clr) {
+      this->pos = pos;
+      this->uv = uv;
+      this->color = clr;
+    }
     vertex() {}
 
     amy::fvec2 pos;
     amy::fvec2 uv;
-    u32 color = 0;
+    ui color = 0;
   };
   class command {
    public:
@@ -57,6 +62,8 @@ class iron {
     drawlist(drawlist&&) noexcept = default;
     drawlist& operator=(drawlist&&) noexcept = default;
 
+    std::vector<command::ref>& data() { return m_data; }
+
     void merge(drawlist* list);
     command::ref newCommand();
     void push(command ::ref cmd);
@@ -65,9 +72,56 @@ class iron {
     void drawSolid();
     void drawTex(texture* tex) { m_tex = tex; }
 
+    /** Draw Api */
+    void drawRect(const fvec2& pos, const fvec2& size, ui color,
+                  int thickness = 1);
+    void drawRectFilled(const fvec2& pos, const fvec2& size, ui color);
+    void drawTriangle(const fvec2& a, const fvec2& b, const fvec2& c, ui color,
+                      int thickness = 1);
+    void drawTriangleFilled(const fvec2& a, const fvec2& b, const fvec2& c,
+                            ui color);
+    void drawCircle(const fvec2& center, float radius, ui color, int segments,
+                    int thickness = 1);
+    void drawCircleFilled(const fvec2& center, float radius, ui color,
+                          int segments);
+    void drawText(const fvec2& pos, const std::string& text, ui color);
+    void drawTextEx(const fvec2& pos, const std::string& text, ui color,
+                    ui flags, const fvec2& box = 0);
+    void drawLine(const fvec2& a, const fvec2& b, ui color, int thickness = 1);
+    void drawPolyLine(const std::vector<fvec2>& points, ui color, ui flags = 0,
+                      int thickness = 1);
+    void drawConvexPolyFilled(const std::vector<fvec2>& points, ui color);
+
+    /** Path api */
+    void pathAdd(const fvec2& pos) { m_path.push_back(std::move(pos)); }
+    void pathClear() { m_path.clear(); }
+    void pathReserve(size_t count) { m_path.reserve(m_path.size() + count); }
+    void pathStroke(ui color, int thickness = 1, ui flags = 0) {
+      drawPolyLine(m_path, color, flags, thickness);
+      pathClear();
+    }
+    void pathFill(ui color) { drawConvexPolyFilled(m_path, color); }
+    void pathArcToN(const fvec2& c, float radius, float a_min, float a_max,
+                    int segments);
+    void pathFastArcToN(const fvec2& c, float radius, float a_min, float a_max,
+                        int segments);
+    void pathRect(const fvec2& a, const fvec2& b, float rounding = 0.f);
+    void pathRectEx(const fvec2& a, const fvec2& b, float rounfing, ui flags);
+
+    void pushClipRect(const fvec4& clip) { clipRects.push(clip); }
+    void popClipRect() {
+      if (!clipRects.empty()) clipRects.pop();
+    }
+
+    operator std::vector<command::ref>&() { return m_data; }
+
    private:
+    void clipCmd(command* ptr);
     std::vector<command::ref> m_data;
-    texture* m_tex;
+    std::vector<fvec2> m_path;
+    texture* m_tex = nullptr;
+    std::stack<fvec4> clipRects;
+    int m_layer = 0;
   };
   iron() = default;
   ~iron() = default;
@@ -76,10 +130,28 @@ class iron {
   static void newFrame();
   static void drawOn(c3d::screen* screen);
   static void draw(const std::vector<command::ref>& data);
+  static texture* whiteTex() { return m_solid; }
+
+  /** Static renderer utility funcs */
+
+  static void rotateCorner(fvec2& pos, float s, float c);
+  static rect primRect(const fvec2& pos, const fvec2& size, float angle = 0.f);
+  static rect primLine(const fvec2& a, const fvec2& b, int thickness = 1);
+  static void cmdQuad(command* cmd, const rect& q, const rect& uv, ui color);
+  static void cmdTriangle(command* cmd, const fvec2& a, const fvec2& b,
+                          const fvec2& c, ui clr);
+  static void cmdConvexPolyFilled(command* cmd,
+                                  const std::vector<fvec2>& points, ui clr,
+                                  texture* tex);
+  static bool inBox(const fvec2& pos, const fvec2& size, const fvec4& area);
+  static bool inBox(const fvec2& pos, const fvec4& area);
+  static bool inBox(const fvec2& a, const fvec2& b, const fvec2& c,
+                    const fvec4& area);
 
  private:
   static void setupShader();
   static void fragConfig();
+  static void initSolidTex();
 
   static std::vector<vertex, linearAllocator<vertex>> m_vbuf;
   static std::vector<u16, linearAllocator<u16>> m_ibuf;
@@ -87,5 +159,6 @@ class iron {
   static c3d::shader* m_shader;
   static mat4 m_mtx;
   static int m_idx, m_vtx;
+  static texture* m_solid;
 };
 }  // namespace amy

include/amethyst/rect.hpp

@@ -0,0 +1,61 @@
+#pragma once
+
+#include <amethyst/types.hpp>
+
+namespace amy {
+class rect {
+ public:
+  rect() : m_top(0), m_bot(0) {}
+  ~rect() = default;
+  rect(const fvec4& t, const fvec4& b) : m_top(t), m_bot(b) {}
+  rect(const fvec2& tl, const fvec2& tr, const fvec2& bl, const fvec2& br)
+      : m_top(tl, tr), m_bot(bl, br) {}
+  rect(const fvec4& duv)
+      : m_top(duv.x, duv.y, duv.z, duv.y), m_bot(duv.x, duv.w, duv.z, duv.w) {}
+
+  fvec2 topLeft() const { return fvec2(m_top.x, m_top.y); }
+  fvec2 topRight() const { return fvec2(m_top.z, m_top.w); }
+  fvec2 botLeft() const { return fvec2(m_bot.x, m_bot.y); }
+  fvec2 botRight() const { return fvec2(m_bot.z, m_bot.w); }
+
+  fvec4 top() const { return m_top; }
+  fvec4& top() { return m_top; }
+  fvec4 bot() const { return m_bot; }
+  fvec4& bot() { return m_bot; }
+
+  rect& topLeft(const fvec2& v) {
+    m_top.x = v.x;
+    m_top.y = v.y;
+    return *this;
+  }
+
+  rect& topRight(const fvec2& v) {
+    m_top.z = v.x;
+    m_top.w = v.y;
+    return *this;
+  }
+
+  rect& botLeft(const fvec2& v) {
+    m_bot.x = v.x;
+    m_bot.y = v.y;
+    return *this;
+  }
+
+  rect& botRight(const fvec2& v) {
+    m_bot.z = v.x;
+    m_bot.w = v.y;
+    return *this;
+  }
+
+  void swapVec2XY() {
+    m_top.SwapXY();
+    m_top.SwapZW();
+    m_bot.SwapXY();
+    m_bot.SwapZW();
+  }
+
+ private:
+  fvec4 m_top;
+  fvec4 m_bot;
+};
+}  // namespace amy

include/amethyst/texture.hpp

@@ -4,6 +4,7 @@
 
 #include <amethyst/asset.hpp>
 #include <amethyst/image.hpp>
+#include <amethyst/rect.hpp>
 #include <amethyst/types.hpp>
 
 namespace amy {
@@ -13,6 +14,8 @@ class texture : public asset {
   texture(cstr& path);
   ~texture();
   void load(cstr& path);
+  void load(const std::vector<uc>& pixels, int w, int h, int bpp = 4,
+            image::format fmt = image::RGBA);
   void unload();
 
   int w() const { return m_size.x; }
@@ -21,6 +24,7 @@ class texture : public asset {
   int& h() { return m_size.y; }
   ivec2 size() const { return m_size; }
   ivec2& size() { return m_size; }
+  rect& uv() { return m_uv; }
 
   C3D_Tex* ptr() { return m_loaded ? &m_tex : nullptr; }
 
@@ -29,6 +33,7 @@ class texture : public asset {
  private:
   C3D_Tex m_tex;
   ivec2 m_size;
+  rect m_uv;
   bool m_loaded = false;
 };
 }  // namespace amy

include/amethyst/types.hpp

@@ -1,10 +1,13 @@
 #pragma once
 
+#include <algorithm>
 #include <amethyst/maths/vec.hpp>
 #include <cinttypes>
 #include <functional>
 #include <map>
 #include <memory>
+#include <numbers>
+#include <stack>
 #include <stdexcept>
 #include <string>
 #include <vector>

source/iron/drawlist.cpp

@@ -1,5 +1,271 @@
 #include <amethyst/iron.hpp>
 
+/** Setup for everything (oder so) */
+enum LiPathRectFlags_ : amy::ui {
+  LiPathRectFlags_None = 0,
+  LiPathRectFlags_KeepTopLeft = (1 << 0),
+  LiPathRectFlags_KeepTopRight = (1 << 1),
+  LiPathRectFlags_KeepBotRight = (1 << 2),
+  LiPathRectFlags_KeepBotLeft = (1 << 3),
+  LiPathRectFlags_KeepTop = (1 << 0) | (1 << 1),
+  LiPathRectFlags_KeepBot = (1 << 2) | (1 << 3),
+  LiPathRectFlags_KeepLeft = (1 << 0) | (1 << 3),
+  LiPathRectFlags_KeepRight = (1 << 1) | (1 << 2),
+};
+
 namespace amy {
-void iron::drawlist::merge(iron::drawlist* list) {}
+constexpr auto __pi = std::numbers::pi;
+
+void iron::drawlist::merge(iron::drawlist* list) {
+  for (size_t i = 0; i < list->m_data.size(); i++) {
+    m_data.push_back(std::move(list->m_data[i]));
+  }
+  list->clear();
+}
+
+void iron::drawlist::clear() { m_data.clear(); }
+
+iron::command::ref iron::drawlist::newCommand() {
+  auto ret = std::make_unique<command>();
+  ret->layer = m_layer;
+  ret->index = m_data.size();
+  ret->tex = m_tex;
+  return ret;
+}
+
+void iron::drawlist::clipCmd(command* ptr) {
+  if (!clipRects.empty()) {
+    ptr->scissorOn = true;
+    ptr->scissorRect = ivec4(clipRects.top());
+  }
+}
+
+void iron::drawlist::push(command::ref cmd) {
+  m_data.push_back(std::move(cmd));
+}
+
+void iron::drawlist::drawSolid() { m_tex = iron::whiteTex(); }
+
+void iron::drawlist::pathArcToN(const fvec2& c, float radius, float a_min,
+                                float a_max, int segments) {
+  // pathAdd(c)
+  pathReserve(segments + 1);
+  for (int i = 0; i < segments; i++) {
+    float a = a_min + ((float)i / (float)segments) * (a_max - a_min);
+    pathAdd(fvec2(c.x + std::cos(a) * radius, c.y + std::sin(a) * radius));
+  }
+}
+
+void iron::drawlist::pathFastArcToN(const fvec2& c, float r, float amin,
+                                    float amax, int s) {
+  /**
+   * Funcion with less division overhead
+   * Usefull for stuff where a lot of calculations are required
+   */
+  float d = (amax - amin) / s;
+  pathReserve(s + 1);
+  for (int i = 0; i <= s; i++) {
+    float a = amin + i * d;
+    pathAdd(fvec2(c.x + std::cos(a) * r, c.y + std::sin(a) * r));
+  }
+}
+
+void iron::drawlist::pathRect(const fvec2& a, const fvec2& b, float rounding) {
+  if (rounding == 0.f) {
+    pathAdd(a);
+    pathAdd(fvec2(b.x, a.y));
+    pathAdd(b);
+    pathAdd(fvec2(a.x, b.y));
+  } else {
+    float r = std::min({rounding, (b.x - a.x) * 0.5f, (b.y - a.y) * 0.5f});
+    /** Calculate Optimal segment count automatically */
+    float corner = __pi * 0.5f;
+    int segments = std::max(3, int(std::ceil(corner / (6.0f * __pi / 180.0f))));
+
+    /**
+     * To Correctly render filled shapes with Paths API
+     * The Commands need to be setup clockwise
+     */
+    /** Top Left */
+    pathAdd(fvec2(a.x + r, a.y));
+    pathFastArcToN(fvec2(b.x - r, a.y + r), r, -__pi / 2.0f, 0.0f, segments);
+    /** Top Right */
+    pathAdd(fvec2(b.x, b.y - r));
+    pathFastArcToN(fvec2(b.x - r, b.y - r), r, 0.0f, __pi / 2.0f, segments);
+    /** Bottom Right */
+    pathAdd(fvec2(a.x + r, b.y));
+    pathFastArcToN(fvec2(a.x + r, b.y - r), r, __pi / 2.0f, __pi, segments);
+    /** Bottom Left */
+    pathAdd(fvec2(a.x, a.y + r));
+    pathFastArcToN(fvec2(a.x + r, a.y + r), r, __pi, 3.0f * __pi / 2.0f,
+                   segments);
+  }
+}
+
+void iron::drawlist::pathRectEx(const fvec2& a, const fvec2& b, float rounding,
+                                ui flags) {
+  if (rounding == 0.f) {
+    pathAdd(a);
+    pathAdd(fvec2(b.x, a.y));
+    pathAdd(b);
+    pathAdd(fvec2(a.x, b.y));
+  } else {
+    float r = std::min({rounding, (b.x - a.x) * 0.5f, (b.y - a.y) * 0.5f});
+    /** Calculate Optimal segment count automatically */
+    float corner = __pi * 0.5f;
+    int segments = std::max(3, int(std::ceil(corner / (6.0f * __pi / 180.0f))));
+
+    /**
+     * To Correctly render filled shapes with Paths API
+     * The Commands need to be setup clockwise
+     */
+    /** Top Left */
+    if (flags & LiPathRectFlags_KeepTopLeft) {
+      pathAdd(a);
+    } else {
+      pathAdd(fvec2(a.x + r, a.y));
+      pathFastArcToN(fvec2(b.x - r, a.y + r), r, -__pi / 2.0f, 0.0f, segments);
+    }
+
+    /** Top Right */
+    if (flags & LiPathRectFlags_KeepTopRight) {
+      pathAdd(fvec2(b.x, a.y));
+    } else {
+      pathAdd(fvec2(b.x, b.y - r));
+      pathFastArcToN(fvec2(b.x - r, b.y - r), r, 0.0f, __pi / 2.0f, segments);
+    }
+    /** Bottom Right */
+    if (flags & LiPathRectFlags_KeepBotRight) {
+      pathAdd(b);
+    } else {
+      pathAdd(fvec2(a.x + r, b.y));
+      pathFastArcToN(fvec2(a.x + r, b.y - r), r, __pi / 2.0f, __pi, segments);
+    }
+    /** Bottom Left */
+    if (flags & LiPathRectFlags_KeepBotLeft) {
+      pathAdd(fvec2(a.x, b.y));
+    } else {
+      pathAdd(fvec2(a.x, a.y + r));
+      pathFastArcToN(fvec2(a.x + r, a.y + r), r, __pi, 3.0f * __pi / 2.0f,
+                     segments);
+    }
+  }
+}
+
+void iron::drawlist::drawRect(const fvec2& pos, const fvec2& size, ui color,
+                              int thickness) {
+  pathRect(pos, pos + size);
+  pathStroke(color, thickness, 1);
+}
+
+void iron::drawlist::drawRectFilled(const fvec2& pos, const fvec2& size,
+                                    ui color) {
+  pathRect(pos, pos + size);
+  pathFill(color);
+}
+
+void iron::drawlist::drawTriangle(const fvec2& a, const fvec2& b,
+                                  const fvec2& c, ui color, int thickness) {
+  pathAdd(a);
+  pathAdd(b);
+  pathAdd(c);
+  pathStroke(color, thickness, 1);
+}
+
+void iron::drawlist::drawTriangleFilled(const fvec2& a, const fvec2& b,
+                                        const fvec2& c, ui color) {
+  pathAdd(a);
+  pathAdd(b);
+  pathAdd(c);
+  pathFill(color);
+}
+
+void iron::drawlist::drawCircle(const fvec2& center, float rad, ui color,
+                                int segments, int thickness) {
+  if (segments <= 0) {
+    // Auto Segment
+  } else {
+    float am = (__pi * 2.0f) * ((float)segments) / (float)segments;
+    pathArcToN(center, rad, 0.f, am, segments);
+  }
+  drawSolid();  // Only Solid Color Supported
+  pathStroke(color, thickness, 1);
+}
+
+void iron::drawlist::drawCircleFilled(const fvec2& center, float rad, ui color,
+                                      int segments) {
+  if (segments <= 0) {
+    // Auto Segment
+  } else {
+    float am = (__pi * 2.0f) * ((float)segments) / (float)segments;
+    pathArcToN(center, rad, 0.f, am, segments);
+  }
+  pathFill(color);
+}
+
+void iron::drawlist::drawPolyLine(const std::vector<fvec2>& points, ui clr,
+                                  ui flags, int thickness) {
+  if (points.size() < 2) {
+    return;
+  }
+  drawSolid();
+  auto cmd = newCommand();
+  bool close = (flags & (1 << 0));
+  int num_points = close ? (int)points.size() : (int)points.size() - 1;
+  if (flags & (1 << 1)) {
+    // TODO: Find a way to draw less garbage looking lines
+  } else {
+    // Non antialiased lines look awful when rendering with thickness != 1
+    for (int i = 0; i < num_points; i++) {
+      int j = (i + 1) == (int)points.size() ? 0 : (i + 1);
+      auto line = primLine(points[i], points[j], thickness);
+      cmdQuad(cmd.get(), line, fvec4(0.f, 1.f, 1.f, 0.f), clr);
+    }
+  }
+  push(std::move(cmd));
+}
+
+void iron::drawlist::drawConvexPolyFilled(const std::vector<fvec2>& points,
+                                          ui clr) {
+  if (points.size() < 3) {
+    return;  // Need at least three points
+  }
+  auto cmd = newCommand();
+  cmdConvexPolyFilled(cmd.get(), points, clr, m_tex);
+  push(std::move(cmd));
+}
+
+void iron::drawlist::drawText(const fvec2& pos, const std::string& text,
+                              ui color) {
+  /*if (!pCurrentFont) {
+    return;
+  }
+  std::vector<Command::Ref> cmds;
+  pCurrentFont->CmdTextEx(cmds, pos, color, pFontScale, text);
+  for (size_t i = 0; i < cmds.size(); i++) {
+    cmds[i]->Index = pDrawList.size();
+    cmds[i]->Layer = Layer;
+    AddCommand(std::move(cmds[i]));
+  }*/
+}
+
+void iron::drawlist::drawTextEx(const fvec2& p, const std::string& text,
+                                ui color, ui flags, const fvec2& box) {
+  /*if (!pCurrentFont) {
+    return;
+  }
+  std::vector<Command::Ref> cmds;
+  pCurrentFont->CmdTextEx(cmds, p, color, pFontScale, text, flags, box);
+  for (size_t i = 0; i < cmds.size(); i++) {
+    cmds[i]->Index = pDrawList.size();
+    cmds[i]->Layer = Layer;
+    AddCommand(std::move(cmds[i]));
+  }*/
+}
+
+void iron::drawlist::drawLine(const fvec2& a, const fvec2& b, ui color, int t) {
+  pathAdd(a);
+  pathAdd(b);
+  pathStroke(color, t);
+}
 }  // namespace amy

source/iron/iron.cpp

@@ -1,4 +1,17 @@
 #include <amethyst/iron.hpp>
+#include <iostream>
+
+#define catch2()                                                        \
+  ;                                                                     \
+  ({                                                                    \
+    std::cout << std::format("{}:{}", __FILE__, __LINE__) << std::endl; \
+    while (aptMainLoop()) {                                             \
+      hidScanInput();                                                   \
+      if (hidKeysDown() & KEY_A) {                                      \
+        break;                                                          \
+      }                                                                 \
+    }                                                                   \
+  })
 
 namespace amy {
 const char* __ironshader__ = R"(; LI7 Shader
@@ -35,17 +48,26 @@ const char* __ironshader__ = R"(; LI7 Shader
 	mov out_color, r1
 	end
 .end)";
+
+// clang-format off
+unsigned char li_shader[] = {
+  0x44, 0x56, 0x4c, 0x42, 0x1, 0x0, 0x0, 0x0, 0xa4, 0x0, 0x0, 0x0, 0x44, 0x56, 0x4c, 0x50, 0x0, 0x0, 0x0, 0x0, 0x28, 0x0, 0x0, 0x0, 0xa, 0x0, 0x0, 0x0, 0x50, 0x0, 0x0, 0x0, 0x9, 0x0, 0x0, 0x0, 0x98, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x4e, 0x1, 0xf0, 0x7, 0x4e, 0x2, 0x8, 0x2, 0x8, 0x3, 0x18, 0x2, 0x8, 0x4, 0x28, 0x2, 0x8, 0x5, 0x38, 0x2, 0x8, 0x6, 0x10, 0x40, 0x4c, 0x7, 0xf1, 0x27, 0x22, 0x8, 0x10, 0x21, 0x4c, 0x0, 0x0, 0x0, 0x88, 0xac, 0x2, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0xa1, 0xa, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x68, 0xc3, 0x6, 0x0, 0x0, 0x0, 0x0, 0x0, 0x64, 0xc3, 0x6, 0x0, 0x0, 0x0, 0x0, 0x0, 0x62, 0xc3, 0x6, 0x0, 0x0, 0x0, 0x0, 0x0, 0x61, 0xc3, 0x6, 0x0, 0x0, 0x0, 0x0, 0x0, 0xaf, 0x2, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x4f, 0xd5, 0x6, 0x0, 0x0, 0x0, 0x0, 0x0, 0x6f, 0x3, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x44, 0x56, 0x4c, 0x45, 0x2, 0x10, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0xa, 0x0, 0x0, 0x0, 0x0, 0x0, 0x7, 0x0, 0x0, 0x0, 0x0, 0x0, 0x40, 0x0, 0x0, 0x0, 0x1, 0x0, 0x0, 0x0, 0x54, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x54, 0x0, 0x0, 0x0, 0x3, 0x0, 0x0, 0x0, 0x6c, 0x0, 0x0, 0x0, 0x1, 0x0, 0x0, 0x0, 0x74, 0x0, 0x0, 0x0, 0xb, 0x0, 0x0, 0x0, 0x2, 0x0, 0x5f, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x3f, 0x0, 0x1, 0x1, 0x37, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0xf, 0x0, 0x0, 0x0, 0x2, 0x0, 0x1, 0x0, 0xf, 0x0, 0x0, 0x0, 0x3, 0x0, 0x2, 0x0, 0xf, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x10, 0x0, 0x13, 0x0, 0x70, 0x72, 0x6f, 0x6a, 0x65, 0x63, 0x74, 0x69, 0x6f, 0x6e, 0x0, 0x0, 
+  };
+// clang-format on
+size_t li_shader_size = 0x124;
 std::vector<iron::vertex, linearAllocator<iron::vertex>> iron::m_vbuf;
 std::vector<u16, linearAllocator<u16>> iron::m_ibuf;
 int iron::uLocProj = 0;
 c3d::shader* iron::m_shader = nullptr;
 mat4 iron::m_mtx;
 int iron::m_idx = 0, iron::m_vtx = 0;
+texture* iron::m_solid = nullptr;
 
 void iron::init() {
   setupShader();
   m_vbuf.resize(4 * 4096);
   m_ibuf.resize(6 * 4096);
+  initSolidTex();
 }
 
 void iron::newFrame() {
@@ -53,11 +75,16 @@ void iron::newFrame() {
   m_vtx = 0;
 }
 
+C3D_Mtx m;
+
 void iron::drawOn(c3d::screen* screen) {
   m_shader->use();
+  Mtx_Identity(&m);
+  Mtx_OrthoTilt(&m, 0.f, (float)screen->width(), (float)screen->height(), 0.f,
+                1.f, -1.f, false);
   m_mtx = mat4::ortho(0.f, (float)screen->width(), (float)screen->height(), 0.f,
                       1.f, -1.f);
-  m_shader->setMat4(uLocProj, m_mtx);
+  m_shader->setMat4(uLocProj, &m);
 }
 
 void iron::draw(const std::vector<iron::command::ref>& data) {
@@ -74,10 +101,12 @@ void iron::draw(const std::vector<iron::command::ref>& data) {
     auto scissorOn = data[i]->scissorOn;
     auto scissor = data[i]->scissorRect;
     auto start = i;
+
     // Loop until a statgechange and copy all data into vertex/index buf
     while (i < data.size() && scissorOn == data[i]->scissorOn &&
            scissor == data[i]->scissorRect && tex == data[i]->tex) {
       auto c = data[i].get();
+
       for (int j = 0; j < c->indexBuf.size(); j++) {
         m_ibuf[m_idx++] = m_vtx + c->indexBuf[i];
       }
@@ -97,7 +126,8 @@ void iron::draw(const std::vector<iron::command::ref>& data) {
 
 void iron::setupShader() {
   m_shader = new c3d::shader();
-  m_shader->compile(__ironshader__);
+  m_shader->load("romfs:/shaders/lithium.shbin");
+  // m_shader->compile(__ironshader__);
   m_shader->input(GPU_FLOAT, 2);          // pos
   m_shader->input(GPU_FLOAT, 2);          // uv
   m_shader->input(GPU_UNSIGNED_BYTE, 4);  // color
@@ -109,4 +139,126 @@ void iron::fragConfig() {
   c3d::frag::src(C3D_Both, GPU_TEXTURE0);
   c3d::frag::func(C3D_Both, GPU_MODULATE);
 }
+
+void iron::initSolidTex() {
+  // i know there is a lot of memory wasted :(
+  std::vector<uc> pixels(16 * 16 * 4, 0xff);
+  m_solid = new texture();
+  m_solid->load(pixels, 16, 16);
+}
+
+bool iron::inBox(const fvec2& pos, const fvec2& size, const fvec4& area) {
+  return (pos.x + size.x >= area.x && pos.y + size.y >= area.y &&
+          pos.x <= area.z && pos.y <= area.w);
+}
+
+bool iron::inBox(const fvec2& pos, const fvec4& area) {
+  return (pos.x > area.x && pos.x < area.x + area.z && pos.y > area.y &&
+          pos.y < area.y + area.w);
+}
+
+bool iron::inBox(const fvec2& a, const fvec2& b, const fvec2& c,
+                 const fvec4& area) {
+  return ((a.x < area.z && b.x < area.z && c.x < area.z) ||
+          (a.y < area.w && b.y < area.w && c.y < area.w) ||
+          (a.x > 0 && b.x > 0 && c.x > 0) || (a.y > 0 && b.y > 0 && c.y > 0));
+}
+
+void iron::rotateCorner(fvec2& pos, float s, float c) {
+  float x = pos.x * c - pos.y * s;
+  float y = pos.y * c - pos.x * s;
+  pos = fvec2(x, y);
+}
+
+rect iron::primRect(const fvec2& pos, const fvec2& size, float angle) {
+  fvec2 c = size * 0.5f;  // Center
+  fvec2 corner[4] = {
+      fvec2(-c.x, -c.y),
+      fvec2(-c.x + size.x, -c.y),
+      fvec2(-c.x, -c.y + size.y),
+      fvec2(-c.x + size.x, -c.y + size.y),
+  };
+
+  // Only rotate if required
+  if (angle != 0.f) {
+    float s = std::sin(angle);
+    float co = std::cos(angle);
+    for (int i = 0; i < 4; i++) {
+      rotateCorner(corner[i], s, co);
+    }
+  }
+
+  // Return Result
+  return rect(corner[0] + pos + c, corner[1] + pos + c, corner[2] + pos + c,
+              corner[3] + pos + c);
+}
+
+rect iron::primLine(const fvec2& a, const fvec2& b, int thickness) {
+  // Using the vec maths api makes the code as short as it is
+  fvec2 dir = a - b;
+  float len = dir.Len();
+  fvec2 unit_dir = dir / len;
+  fvec2 perpendicular(-unit_dir.y, unit_dir.x);
+  fvec2 off = perpendicular * ((float)thickness * 0.5f);
+
+  return rect(a + off, b + off, a - off, b - off);
+}
+
+void iron::cmdQuad(command* cmd, const rect& q, const rect& uv, ui color) {
+  cmd->add(0).add(1).add(2);
+  cmd->add(0).add(2).add(3);
+  cmd->add(vertex(q.botRight(), uv.botRight(), color));
+  cmd->add(vertex(q.topRight(), uv.topRight(), color));
+  cmd->add(vertex(q.topLeft(), uv.topLeft(), color));
+  cmd->add(vertex(q.botLeft(), uv.botLeft(), color));
+}
+
+void iron::cmdTriangle(command* cmd, const fvec2& a, const fvec2& b,
+                       const fvec2& c, ui color) {
+  cmd->add(2).add(1).add(0);  // reverse cause otherwise invisible
+  cmd->add(vertex(a, fvec2(0, 1), color));
+  cmd->add(vertex(b, fvec2(1, 1), color));
+  cmd->add(vertex(c, fvec2(1, 0), color));
+}
+
+void iron::cmdConvexPolyFilled(command* cmd, const std::vector<fvec2>& points,
+                               ui color, texture* tex) {
+  if (points.size() < 3 || tex == nullptr) {
+#ifdef AMY_GOD_DEV
+    return;
+#else
+    throw std::runtime_error("[amy] iron: trying to render convex poly with " +
+                             std::to_string(points.size()) +
+                             " points and texture " + std::to_string((ui)tex));
+#endif
+  }
+  // Support for Custom Textures (UV calculation)
+  float minX = points[0].x, minY = points[0].y;
+  float maxX = minX, maxY = minY;
+  // Check for the max and min Positions
+  for (const auto& it : points) {
+    if (it.x < minX) minX = it.x;
+    if (it.y < minY) minY = it.y;
+    if (it.x > maxX) maxX = it.x;
+    if (it.y > maxY) maxY = it.y;
+  }
+  // Get Short defines for UV
+  // (Bottom Right is not required)
+  auto uv_tl = tex->uv().topLeft();
+  auto uv_tr = tex->uv().topRight();
+  auto uv_bl = tex->uv().botLeft();
+
+  // Render
+  for (int i = 2; i < (int)points.size(); i++) {
+    cmd->add(0).add(i).add(i - 1);
+  }
+  for (int i = 0; i < (int)points.size(); i++) {
+    // Calculate U and V coords
+    float u =
+        uv_tl.x + ((points[i].x - minX) / (maxX - minX)) * (uv_tr.x - uv_tl.x);
+    float v =
+        uv_tl.y + ((points[i].y - minY) / (maxY - minY)) * (uv_bl.y - uv_tl.y);
+    cmd->add(vertex(points[i], fvec2(u, v), color));
+  }
+}
 }  // namespace amy

source/texture.cpp

@@ -44,35 +44,42 @@ void texture::unload() {
 }
 
 void texture::load(cstr& path) {
-  unload();
   image img(path);
   if (img.width() > 1024 || img.height() > 1024) {
     throw std::runtime_error("Max Texture Size is 1024x1024!");
   }
+  load(img.getBuffer(), img.width(), img.height(), img.bpp(), img.fmt());
+}
 
-  int bpp = img.bpp();
+void texture::load(const std::vector<uc>& pixels, int w, int h, int bpp,
-  m_size.x = img.width();
+                   image::format fmt) {
+  if (w > 1024 || h > 1024) {
+    throw std::runtime_error("Max Texture Size is 1024x1024!");
+  }
+  unload();
+
+  m_size.x = w;
   if (utils::isSingleBitNum(m_size.x)) {
     m_size.x = utils::nextPow2(m_size.x);
   }
-  m_size.y = img.height();
+  m_size.y = h;
   if (utils::isSingleBitNum(m_size.y)) {
     m_size.y = utils::nextPow2(m_size.y);
   }
   auto filter = GPU_NEAREST;
-  auto format = image2TexFmt(img.fmt());
+  auto format = image2TexFmt(fmt);
   C3D_TexInit(&m_tex, (u16)m_size.x, (u16)m_size.y, format);
   C3D_TexSetFilter(&m_tex, filter, filter);
   // Using std::fill_n instead cause i hate this error lines
   // under the memset func in my editor
   std::fill_n((unsigned char*)m_tex.data, m_tex.size, 0);
-  for (int x = 0; x < img.width(); x++) {
+  for (int x = 0; x < w; x++) {
-    for (int y = 0; y < img.height(); y++) {
+    for (int y = 0; y < h; y++) {
       int dst_pos = tile3dsTex(x, y, m_size.x) * bpp;
-      int src_pos = (y * img.width() + x) * bpp;
+      int src_pos = (y * w + x) * bpp;
       /// Best idea i had
       for (int i = 0; i < bpp; i++) {
-        ((u8*)m_tex.data)[dst_pos + bpp - 1 - i] = img[src_pos + i];
+        ((u8*)m_tex.data)[dst_pos + bpp - 1 - i] = pixels[src_pos + i];
       }
     }
   }