wip: uncommitted changes from claude run on issue #3

examples/RTStress/closesthit.wgsl

@@ -0,0 +1,54 @@
+// RTStress closest-hit (runs in SHADE). Computes flat-shaded Lambert from
+// the hit triangle's geometric normal, accumulates ambient, and — if the
+// surface faces the sun — emits a shadow ray toward the sun. The shadow
+// ray's miss (sun visible) adds the direct term; its hit (occluded) adds
+// nothing because RT_FLAG_SKIP_CLOSEST_HIT suppresses closesthit on hit.
+//
+// Payload declared here so the assembler sees it before wfPayload / SHADE.
+struct Payload {
+    color:     vec3<f32>,   // shadow ray: pending direct contribution
+    shadowRay: u32,         // 0 primary, 1 shadow
+};
+
+const SUN_DIR_TO_LIGHT: vec3<f32> = vec3<f32>(0.40, 0.85, 0.35);
+const SUN_COLOR:        vec3<f32> = vec3<f32>(1.15, 1.05, 0.90);
+const AMBIENT_COLOR:    vec3<f32> = vec3<f32>(0.12, 0.13, 0.18);
+
+// Cheap per-instance albedo so the grid reads as distinct cubes (and any
+// TLAS flicker as instance count scales is obvious).
+fn instanceAlbedo(i: u32) -> vec3<f32> {
+    let h = i * 2654435761u;
+    return vec3<f32>(
+        0.35 + 0.6 * f32((h >>  0u) & 255u) / 255.0,
+        0.35 + 0.6 * f32((h >>  8u) & 255u) / 255.0,
+        0.35 + 0.6 * f32((h >> 16u) & 255u) / 255.0);
+}
+
+fn closesthit_main(ray: RayDesc, hit: HitInfo, payload: ptr<function, Payload>) {
+    let meshRec = meshRecords[tlasEntries[hit.instanceId].blasMeshIdx];
+    let verts = _rtFetchTri(meshRec, hit.primitiveId);
+    let nObj = normalize(cross(verts[1] - verts[0], verts[2] - verts[0]));
+    let nWorld = normalize(vec3<f32>(
+        dot(hit.objectToWorldR0.xyz, nObj),
+        dot(hit.objectToWorldR1.xyz, nObj),
+        dot(hit.objectToWorldR2.xyz, nObj)));
+
+    let albedo   = instanceAlbedo(hit.customIndex);
+    let worldPos = ray.origin + ray.direction * hit.t;
+    let viewDir  = -ray.direction;
+    let nFacing  = select(-nWorld, nWorld, dot(nWorld, viewDir) > 0.0);
+    let sunDir   = normalize(SUN_DIR_TO_LIGHT);
+    let nDotL    = max(0.0, dot(nFacing, sunDir));
+
+    rtAccumulate(albedo * AMBIENT_COLOR);
+
+    if (nDotL > 0.0) {
+        var sp: Payload;
+        sp.color     = albedo * SUN_COLOR * nDotL;
+        sp.shadowRay = 1u;
+        let shadowOrigin = worldPos + nFacing * 0.05;
+        rtEmitRay(shadowOrigin, 0.01, sunDir, 100000.0,
+                  RT_FLAG_SKIP_CLOSEST_HIT | RT_FLAG_TERMINATE_ON_FIRST_HIT,
+                  0xFFu, 0u, 0u, sp);
+    }
+}

examples/RTStress/main.cpp

@@ -0,0 +1,200 @@
+// RTStress — the standing many-instance wavefront RT benchmark. An
+// N×N×N grid of a small cube mesh (one BLAS, many TLAS instances), shaded
+// with primary + shadow rays through the wavefront pipeline. The grid edge
+// `kGrid` is the instance-count knob: 8 → 512, 16 → 4096, 20 → 8000
+// (LBVH_MAX = 16384). Frame time is printed to the console each second so
+// fps-vs-instance-count can be read off without external tooling; the JS
+// bridge additionally prints a GPU timestamp-query per-pass breakdown.
+//
+// WebGPU/DOM only — the wavefront tracer is the WebGPU software RT path.
+
+#ifndef CRAFTER_GRAPHICS_WINDOW_DOM
+int main() { return 0; }   // native path is hardware RT; out of scope here
+#else
+
+import Crafter.Graphics;
+import Crafter.Math;
+import Crafter.Event;
+import std;
+
+using namespace Crafter;
+namespace fs = std::filesystem;
+
+namespace {
+    // Instance-count knob. instances = kGrid³. Bump to 16 (4096) or 20
+    // (8000) to stress the TLAS; the LBVH build caps at 16384.
+    constexpr int   kGrid    = 8;
+    constexpr float kSpacing = 2.5f;
+    constexpr float kHalf    = 0.5f;   // cube half-extent
+
+    struct CameraGPU {
+        float origin[3];   float pad0;
+        float right[3];    float tanHalf;
+        float up[3];       float aspect;
+        float forward[3];  float pad1;
+    };
+    static_assert(sizeof(CameraGPU) == 64);
+}
+
+int main() {
+    const int instanceCount = kGrid * kGrid * kGrid;
+    std::println("[RTStress] grid {}^3 = {} instances", kGrid, instanceCount);
+
+    Device::Initialize();
+    static Window window(1280, 720, "RTStress");
+    auto cmd = window.StartInit();
+
+    DescriptorHeapWebGPU heap;
+    heap.Initialize(/*images*/ 1, /*buffers*/ 2, /*samplers*/ 1);
+
+    std::array<WebGPUShader, 4> shaders {{
+        WebGPUShader(fs::path("raygen.wgsl"),     "raygen_main",     WebGPURTStage::Raygen),
+        WebGPUShader(fs::path("miss.wgsl"),       "miss_main",       WebGPURTStage::Miss),
+        WebGPUShader(fs::path("closesthit.wgsl"), "closesthit_main", WebGPURTStage::ClosestHit),
+        WebGPUShader(fs::path("resolve.wgsl"),    "resolve_main",    WebGPURTStage::Resolve),
+    }};
+    ShaderBindingTableWebGPU sbt;
+    sbt.Init(shaders);
+
+    std::array<RTShaderGroup, 1> raygenGroups {{ { .type = RTShaderGroupType::General,           .generalShader = 0 } }};
+    std::array<RTShaderGroup, 1> missGroups   {{ { .type = RTShaderGroupType::General,           .generalShader = 1 } }};
+    std::array<RTShaderGroup, 1> hitGroups    {{ { .type = RTShaderGroupType::TrianglesHitGroup, .closestHitShader = 2 } }};
+
+    // One user binding: the camera storage buffer at @group(3).
+    std::array<UICustomBinding, 1> bindings {{
+        { .group = 3, .binding = 0, .kind = UICustomBindingKind::Buffer, ._pad = 0, .pushOffset = 0 },
+    }};
+
+    PipelineRTWebGPU pipeline;
+    pipeline.Init(cmd, raygenGroups, missGroups, hitGroups, sbt, bindings);
+
+    // ── Unit cube mesh (8 verts, 12 tris). ────────────────────────────
+    static std::array<Vector<float, 3, 3>, 8> verts {{
+        {-kHalf, -kHalf, -kHalf}, { kHalf, -kHalf, -kHalf},
+        { kHalf,  kHalf, -kHalf}, {-kHalf,  kHalf, -kHalf},
+        {-kHalf, -kHalf,  kHalf}, { kHalf, -kHalf,  kHalf},
+        { kHalf,  kHalf,  kHalf}, {-kHalf,  kHalf,  kHalf},
+    }};
+    static std::array<std::uint32_t, 36> indices {{
+        0,1,2, 0,2,3,   5,4,7, 5,7,6,   4,0,3, 4,3,7,
+        1,5,6, 1,6,2,   4,5,1, 4,1,0,   3,2,6, 3,6,7,
+    }};
+    static Mesh cube;
+    cube.Build(verts, indices, cmd);
+
+    // ── Camera buffer + handle array. ─────────────────────────────────
+    WebGPUBuffer<CameraGPU, true> cameraBuf;
+    cameraBuf.Create(1);
+    static std::array<std::uint32_t, 1> userHandles { cameraBuf.handle };
+
+    // ── Instance grid. Reserve so RenderingElement3D::Add pointers stay
+    //    valid across vector growth. ─────────────────────────────────────
+    static std::vector<RenderingElement3D> renderers;
+    renderers.reserve(static_cast<std::size_t>(instanceCount));
+    const float origin0 = -0.5f * static_cast<float>(kGrid - 1) * kSpacing;
+    for (int x = 0; x < kGrid; ++x)
+    for (int y = 0; y < kGrid; ++y)
+    for (int z = 0; z < kGrid; ++z) {
+        renderers.emplace_back();
+        RenderingElement3D& r = renderers.back();
+        auto& tx = r.instance.transform.matrix;
+        tx[0][0] = 1; tx[0][1] = 0; tx[0][2] = 0; tx[0][3] = origin0 + float(x) * kSpacing;
+        tx[1][0] = 0; tx[1][1] = 1; tx[1][2] = 0; tx[1][3] = origin0 + float(y) * kSpacing;
+        tx[2][0] = 0; tx[2][1] = 0; tx[2][2] = 1; tx[2][3] = origin0 + float(z) * kSpacing;
+        r.instance.instanceCustomIndex                    = static_cast<std::uint32_t>(renderers.size() - 1);
+        r.instance.mask                                   = 0xFF;
+        r.instance.instanceShaderBindingTableRecordOffset = 0;
+        r.instance.flags                                  = kRTGeometryInstanceForceOpaque;
+        r.instance.accelerationStructureReference         = cube.blasAddr;
+        RenderingElement3D::Add(&r);
+    }
+    RenderingElement3D::BuildTLAS(cmd, 0);
+
+    window.descriptorHeap = &heap;
+    window.FinishInit();
+
+    RTPass rtPass(&pipeline);
+    rtPass.handlesPtr   = userHandles.data();
+    rtPass.handlesCount = static_cast<std::uint32_t>(userHandles.size());
+    rtPass.maxDepth     = 2;   // primary + shadow
+    window.passes.push_back(&rtPass);
+
+    // ── Free camera framing the grid from a corner. ───────────────────
+    const float ext = float(kGrid - 1) * kSpacing;
+    struct CamState {
+        Vector<float, 3, 4> position;
+        float yaw;
+        float pitch;
+    } cam {
+        Vector<float, 3, 4>{ ext * 1.4f, ext * 1.0f, ext * 1.4f },
+        0.0f, 0.0f,
+    };
+    {
+        // Aim at the grid centre (origin).
+        Vector<float, 3, 4> d { -cam.position.x, -cam.position.y, -cam.position.z };
+        const float len = std::sqrt(d.x*d.x + d.y*d.y + d.z*d.z);
+        cam.yaw   = std::atan2(d.z, d.x);
+        cam.pitch = std::asin(d.y / len);
+    }
+
+    Input::Map inputMap;
+    Input::Action& moveAct = inputMap.AddAction("Move", Input::ActionType::Vector2);
+    Input::Action& lookAct = inputMap.AddAction("Look", Input::ActionType::Vector2);
+    moveAct.bindings = { Input::WASDBind{
+        Key(CrafterKeys::W), Key(CrafterKeys::S), Key(CrafterKeys::A), Key(CrafterKeys::D) } };
+    lookAct.bindings = { Input::MouseDeltaBind{ 1.0f } };
+    inputMap.Attach(window);
+
+    const float kMoveSpeed = ext * 0.8f;
+    const float kLookSens  = 0.05f;
+    const float kDt        = 1.0f / 60.0f;
+
+    static int   frames = 0;
+    static double tAccum = 0.0;
+    EventListener<void> camTick(&window.onBeforeUpdate, [&]() {
+        inputMap.Tick();
+        cam.yaw   += lookAct.vector2.x * kLookSens;
+        cam.pitch -= lookAct.vector2.y * kLookSens;
+        cam.pitch  = std::clamp(cam.pitch, -1.55f, 1.55f);
+
+        const float cp = std::cos(cam.pitch), sp = std::sin(cam.pitch);
+        const float cy = std::cos(cam.yaw),   sy = std::sin(cam.yaw);
+        Vector<float, 3, 4> forward { cp * cy, sp, cp * sy };
+        Vector<float, 3, 4> worldUp { 0.0f, 1.0f, 0.0f };
+        Vector<float, 3, 4> right { forward.y*worldUp.z - forward.z*worldUp.y,
+                                    forward.z*worldUp.x - forward.x*worldUp.z,
+                                    forward.x*worldUp.y - forward.y*worldUp.x };
+        const float rLen = std::sqrt(right.x*right.x + right.y*right.y + right.z*right.z);
+        right.x /= rLen; right.y /= rLen; right.z /= rLen;
+        Vector<float, 3, 4> up { right.y*forward.z - right.z*forward.y,
+                                 right.z*forward.x - right.x*forward.z,
+                                 right.x*forward.y - right.y*forward.x };
+
+        const float dx = moveAct.vector2.x * kMoveSpeed * kDt;
+        const float dy = moveAct.vector2.y * kMoveSpeed * kDt;
+        cam.position.x += right.x*dx + forward.x*dy;
+        cam.position.y += right.y*dx + forward.y*dy;
+        cam.position.z += right.z*dx + forward.z*dy;
+
+        CameraGPU& g = cameraBuf.value[0];
+        g.origin[0]=cam.position.x; g.origin[1]=cam.position.y; g.origin[2]=cam.position.z; g.pad0=0;
+        g.right[0]=right.x;   g.right[1]=right.y;   g.right[2]=right.z;
+        g.up[0]=up.x;         g.up[1]=up.y;         g.up[2]=up.z;
+        g.forward[0]=forward.x; g.forward[1]=forward.y; g.forward[2]=forward.z;
+        g.aspect  = float(window.width) / float(window.height);
+        g.tanHalf = std::tan(70.0f * 3.14159265f / 360.0f);
+        g.pad1    = 0;
+        cameraBuf.FlushDevice();
+
+        if (++frames >= 60) {
+            std::println("[RTStress] {} instances @ ~{} frames since last report", instanceCount, frames);
+            frames = 0;
+        }
+    });
+
+    window.Render();
+    window.StartUpdate();
+    window.StartSync();
+    return 0;
+}
+#endif

examples/RTStress/miss.wgsl

@@ -0,0 +1,11 @@
+// RTStress miss (runs in SHADE). Primary miss → sky gradient. Shadow miss
+// → the sun is unoccluded, so add the pending direct contribution.
+fn miss_main(ray: RayDesc, payload: ptr<function, Payload>) {
+    if ((*payload).shadowRay == 1u) {
+        rtAccumulate((*payload).color);
+        return;
+    }
+    let t = clamp(ray.direction.y * 0.5 + 0.5, 0.0, 1.0);
+    rtAccumulate(mix(vec3<f32>(0.50, 0.62, 0.85),
+                     vec3<f32>(0.90, 0.94, 1.00), t));
+}

examples/RTStress/project.cpp

@@ -0,0 +1,46 @@
+import std;
+import Crafter.Build;
+namespace fs = std::filesystem;
+using namespace Crafter;
+
+extern "C" Configuration CrafterBuildProject(std::span<const std::string_view> args) {
+    bool isWasm = false;
+    for (std::string_view a : args) {
+        if (a.starts_with("--target=") && a.find("wasm") != std::string_view::npos) {
+            isWasm = true;
+            break;
+        }
+    }
+
+    std::vector<std::string> graphicsArgs(args.begin(), args.end());
+    Configuration* graphics = LocalProject({
+        .projectFile = "../../project.cpp",
+        .args = graphicsArgs,
+    });
+
+    Configuration cfg;
+    cfg.path = "./";
+    cfg.name = "RTStress";
+    cfg.outputName = "RTStress";
+    cfg.type = ConfigurationType::Executable;
+    if (isWasm) {
+        cfg.target = "wasm32-wasip1";
+        cfg.defines.push_back({"CRAFTER_GRAPHICS_WINDOW_DOM", ""});
+        cfg.compileFlags.push_back("-msimd128");
+    }
+    ApplyStandardArgs(cfg, args);
+    cfg.dependencies = { graphics };
+
+    std::array<fs::path, 0> ifaces = {};
+    std::array<fs::path, 1> impls = { "main" };
+    cfg.GetInterfacesAndImplementations(ifaces, impls);
+
+    if (isWasm) {
+        cfg.files.emplace_back(fs::path("raygen.wgsl"));
+        cfg.files.emplace_back(fs::path("closesthit.wgsl"));
+        cfg.files.emplace_back(fs::path("miss.wgsl"));
+        cfg.files.emplace_back(fs::path("resolve.wgsl"));
+        EnableWasiBrowserRuntime(cfg);
+    }
+    return cfg;
+}

examples/RTStress/raygen.wgsl

@@ -0,0 +1,35 @@
+// RTStress raygen (runs in GENERATE). Host-driven pinhole camera at
+// @group(3) (groups 0..2 are reserved by the wavefront pipeline:
+// 0 = WfParams, 1 = data heaps, 2 = indirect args).
+struct Camera {
+    origin:  vec3<f32>,
+    pad0:    f32,
+    right:   vec3<f32>,
+    tanHalf: f32,
+    up:      vec3<f32>,
+    aspect:  f32,
+    forward: vec3<f32>,
+    pad1:    f32,
+};
+@group(3) @binding(0) var<storage, read> camera : Camera;
+
+fn raygen_main(gid: vec3<u32>) {
+    if (gid.x >= wfParams.surfaceW || gid.y >= wfParams.surfaceH) { return; }
+
+    let pixelf = vec2<f32>(f32(gid.x), f32(gid.y));
+    let res    = vec2<f32>(f32(wfParams.surfaceW), f32(wfParams.surfaceH));
+    let uv     = (pixelf + vec2<f32>(0.5)) / res;
+    let ndc    = uv * 2.0 - vec2<f32>(1.0);
+
+    let direction = normalize(
+        camera.right   * (ndc.x  * camera.aspect * camera.tanHalf) +
+        camera.up      * (-ndc.y * camera.tanHalf) +
+        camera.forward);
+
+    var p: Payload;
+    p.color     = vec3<f32>(0.0);
+    p.shadowRay = 0u;
+
+    rtEmitPrimaryRay(camera.origin, 0.01, direction, 100000.0,
+                     0u, 0xFFu, 0u, 0u, p);
+}

examples/RTStress/resolve.wgsl

@@ -0,0 +1,7 @@
+// RTStress RESOLVE-stage tonemap: Reinhard + gamma 2.2 over the linear
+// accumulator. Registered as a WebGPURTStage::Resolve shader.
+fn resolve_main(coord: vec2<u32>, hdr: vec4<f32>) -> vec4<f32> {
+    let mapped = hdr.rgb / (hdr.rgb + vec3<f32>(1.0));
+    let g      = pow(mapped, vec3<f32>(1.0 / 2.2));
+    return vec4<f32>(g, 1.0);
+}