fix(gpu): Resolve high-DPI squished rendering and 3D shadow bugs

- Implemented dynamic resolution support in all shaders and effects.
- Added explicit viewport setting for all render passes to ensure correct scaling.
- Fixed 3D shadow mapping by adding PLANE support and standardizing soft shadow logic.
- Propagated resize events through the Effect hierarchy.
- Applied project-wide code formatting.

1 files changed, 62 insertions, 36 deletions

diff --git a/src/3d/visual_debug.cc b/src/3d/visual_debug.cc
index 361372b..daa1033 100644
--- a/src/3d/visual_debug.cc
+++ b/src/3d/visual_debug.cc
@@ -5,8 +5,8 @@
 
 #if !defined(STRIP_ALL)
 
-#include <cstring>
 #include <cstdio>
+#include <cstring>
 
 // Simple shader for drawing colored lines
 static const char* kDebugShaderCode = R"(
@@ -42,10 +42,11 @@ fn fs_main(in : VertexOutput) -> @location(0) vec4<f32> {
 void VisualDebug::init(WGPUDevice device, WGPUTextureFormat format) {
   device_ = device;
   create_pipeline(format);
-  
+
   // Initial capacity for vertex buffer (e.g., 1024 lines)
-  vertex_buffer_capacity_ = 1024 * 2 * sizeof(float) * 6; // 2 verts per line, 6 floats per vert
-  
+  vertex_buffer_capacity_ =
+      1024 * 2 * sizeof(float) * 6; // 2 verts per line, 6 floats per vert
+
   WGPUBufferDescriptor vb_desc = {};
   vb_desc.usage = WGPUBufferUsage_Vertex | WGPUBufferUsage_CopyDst;
   vb_desc.size = vertex_buffer_capacity_;
@@ -58,11 +59,16 @@ void VisualDebug::init(WGPUDevice device, WGPUTextureFormat format) {
 }
 
 void VisualDebug::shutdown() {
-  if (pipeline_) wgpuRenderPipelineRelease(pipeline_);
-  if (bind_group_layout_) wgpuBindGroupLayoutRelease(bind_group_layout_);
-  if (vertex_buffer_) wgpuBufferRelease(vertex_buffer_);
-  if (uniform_buffer_) wgpuBufferRelease(uniform_buffer_);
-  if (bind_group_) wgpuBindGroupRelease(bind_group_);
+  if (pipeline_)
+    wgpuRenderPipelineRelease(pipeline_);
+  if (bind_group_layout_)
+    wgpuBindGroupLayoutRelease(bind_group_layout_);
+  if (vertex_buffer_)
+    wgpuBufferRelease(vertex_buffer_);
+  if (uniform_buffer_)
+    wgpuBufferRelease(uniform_buffer_);
+  if (bind_group_)
+    wgpuBindGroupRelease(bind_group_);
 }
 
 void VisualDebug::create_pipeline(WGPUTextureFormat format) {
@@ -82,7 +88,8 @@ void VisualDebug::create_pipeline(WGPUTextureFormat format) {
   WGPUPipelineLayoutDescriptor pl_desc = {};
   pl_desc.bindGroupLayoutCount = 1;
   pl_desc.bindGroupLayouts = &bind_group_layout_;
-  WGPUPipelineLayout pipeline_layout = wgpuDeviceCreatePipelineLayout(device_, &pl_desc);
+  WGPUPipelineLayout pipeline_layout =
+      wgpuDeviceCreatePipelineLayout(device_, &pl_desc);
 
   // Shader
 #if defined(DEMO_CROSS_COMPILE_WIN32)
@@ -98,7 +105,8 @@ void VisualDebug::create_pipeline(WGPUTextureFormat format) {
   WGPUShaderModuleDescriptor shader_desc = {};
   shader_desc.nextInChain = (const WGPUChainedStruct*)&wgsl_desc.chain;
 #endif
-  WGPUShaderModule shader_module = wgpuDeviceCreateShaderModule(device_, &shader_desc);
+  WGPUShaderModule shader_module =
+      wgpuDeviceCreateShaderModule(device_, &shader_desc);
 
   // Vertex State
   WGPUVertexAttribute attributes[2];
@@ -137,7 +145,7 @@ void VisualDebug::create_pipeline(WGPUTextureFormat format) {
   fragment_state.entryPoint = {"fs_main", 7};
 #endif
   fragment_state.targetCount = 1;
-  
+
   WGPUColorTargetState color_target = {};
   color_target.format = format;
   color_target.writeMask = WGPUColorWriteMask_All;
@@ -152,7 +160,8 @@ void VisualDebug::create_pipeline(WGPUTextureFormat format) {
   WGPUDepthStencilState depth_stencil = {};
   depth_stencil.format = WGPUTextureFormat_Depth24Plus;
   depth_stencil.depthWriteEnabled = WGPUOptionalBool_False; // Don't write depth
-  depth_stencil.depthCompare = WGPUCompareFunction_Less; // But do test against it
+  depth_stencil.depthCompare =
+      WGPUCompareFunction_Less; // But do test against it
   pipeline_desc.depthStencil = &depth_stencil;
 
   pipeline_desc.multisample.count = 1;
@@ -177,9 +186,12 @@ void VisualDebug::add_box(const vec3& c, const vec3& e, const vec3& color) {
 
   // 12 edges (each 2 vertices)
   DebugLine edges[] = {
-      {p0, p1, color}, {p1, p2, color}, {p2, p3, color}, {p3, p0, color}, // Front face
-      {p4, p5, color}, {p5, p6, color}, {p6, p7, color}, {p7, p4, color}, // Back face
-      {p0, p4, color}, {p1, p5, color}, {p2, p6, color}, {p3, p7, color}  // Connecting edges
+      {p0, p1, color}, {p1, p2, color},
+      {p2, p3, color}, {p3, p0, color}, // Front face
+      {p4, p5, color}, {p5, p6, color},
+      {p6, p7, color}, {p7, p4, color}, // Back face
+      {p0, p4, color}, {p1, p5, color},
+      {p2, p6, color}, {p3, p7, color} // Connecting edges
   };
 
   for (const auto& l : edges) {
@@ -189,7 +201,8 @@ void VisualDebug::add_box(const vec3& c, const vec3& e, const vec3& color) {
 
 void VisualDebug::update_buffers(const mat4& view_proj) {
   // Update Uniforms
-  wgpuQueueWriteBuffer(wgpuDeviceGetQueue(device_), uniform_buffer_, 0, &view_proj, sizeof(mat4));
+  wgpuQueueWriteBuffer(wgpuDeviceGetQueue(device_), uniform_buffer_, 0,
+                       &view_proj, sizeof(mat4));
 
   // Update Vertices
   size_t required_size = lines_.size() * 2 * sizeof(float) * 6;
@@ -207,38 +220,51 @@ void VisualDebug::update_buffers(const mat4& view_proj) {
     std::vector<float> vertex_data;
     vertex_data.reserve(lines_.size() * 12); // 2 verts * 6 floats
     for (const auto& line : lines_) {
-      vertex_data.push_back(line.start.x); vertex_data.push_back(line.start.y); vertex_data.push_back(line.start.z);
-      vertex_data.push_back(line.color.x); vertex_data.push_back(line.color.y); vertex_data.push_back(line.color.z);
-      
-      vertex_data.push_back(line.end.x); vertex_data.push_back(line.end.y); vertex_data.push_back(line.end.z);
-      vertex_data.push_back(line.color.x); vertex_data.push_back(line.color.y); vertex_data.push_back(line.color.z);
+      vertex_data.push_back(line.start.x);
+      vertex_data.push_back(line.start.y);
+      vertex_data.push_back(line.start.z);
+      vertex_data.push_back(line.color.x);
+      vertex_data.push_back(line.color.y);
+      vertex_data.push_back(line.color.z);
+
+      vertex_data.push_back(line.end.x);
+      vertex_data.push_back(line.end.y);
+      vertex_data.push_back(line.end.z);
+      vertex_data.push_back(line.color.x);
+      vertex_data.push_back(line.color.y);
+      vertex_data.push_back(line.color.z);
     }
-    wgpuQueueWriteBuffer(wgpuDeviceGetQueue(device_), vertex_buffer_, 0, vertex_data.data(), vertex_data.size() * sizeof(float));
+    wgpuQueueWriteBuffer(wgpuDeviceGetQueue(device_), vertex_buffer_, 0,
+                         vertex_data.data(),
+                         vertex_data.size() * sizeof(float));
   }
-  
-  // Re-create bind group if needed (e.g. if uniform buffer changed, though here it's static)
+
+  // Re-create bind group if needed (e.g. if uniform buffer changed, though here
+  // it's static)
   if (!bind_group_) {
-      WGPUBindGroupEntry bg_entry = {};
-      bg_entry.binding = 0;
-      bg_entry.buffer = uniform_buffer_;
-      bg_entry.size = sizeof(mat4);
+    WGPUBindGroupEntry bg_entry = {};
+    bg_entry.binding = 0;
+    bg_entry.buffer = uniform_buffer_;
+    bg_entry.size = sizeof(mat4);
 
-      WGPUBindGroupDescriptor bg_desc = {};
-      bg_desc.layout = bind_group_layout_;
-      bg_desc.entryCount = 1;
-      bg_desc.entries = &bg_entry;
-      bind_group_ = wgpuDeviceCreateBindGroup(device_, &bg_desc);
+    WGPUBindGroupDescriptor bg_desc = {};
+    bg_desc.layout = bind_group_layout_;
+    bg_desc.entryCount = 1;
+    bg_desc.entries = &bg_entry;
+    bind_group_ = wgpuDeviceCreateBindGroup(device_, &bg_desc);
   }
 }
 
 void VisualDebug::render(WGPURenderPassEncoder pass, const mat4& view_proj) {
-  if (lines_.empty()) return;
+  if (lines_.empty())
+    return;
 
   update_buffers(view_proj);
 
   wgpuRenderPassEncoderSetPipeline(pass, pipeline_);
   wgpuRenderPassEncoderSetBindGroup(pass, 0, bind_group_, 0, nullptr);
-  wgpuRenderPassEncoderSetVertexBuffer(pass, 0, vertex_buffer_, 0, lines_.size() * 2 * sizeof(float) * 6);
+  wgpuRenderPassEncoderSetVertexBuffer(pass, 0, vertex_buffer_, 0,
+                                       lines_.size() * 2 * sizeof(float) * 6);
   wgpuRenderPassEncoderDraw(pass, (uint32_t)lines_.size() * 2, 1, 0, 0);
 
   lines_.clear(); // Clear for next frame