perf(compositor): reuse per-element rasters in-place — take-16 GC-churn fix

Slice 10 from today-slices (d1126dd): the paste cache minted a fresh raster array per new source frame (webcam ~30 keys/s ≈ 18MB/s LOH churn → gen2 pauses that ate camera frames). Fix: _elementRasters — re-rasterize INTO the element's existing array when geometry holds, so steady-state allocation ≈ 0. The paste cache keys on (array+epoch) still protect correctness; _elementRasters[element] tracks the element's current raster so stale keys can never paste a mid-update array. MaxPasteEntries 48→256 (was a leak guard, not an allocation stream). Regression test: PasteCache_RingRecycledSources_ReRasterInPlace_WithoutAllocationChurn. Build 0 warnings, 289 tests.
2026-09-05 16:47:49 -07:00
parent 87509bcf99
commit 6d11e8b3fc
2 changed files with 81 additions and 5 deletions
@@ -25,8 +25,9 @@ public sealed class SceneCompositor
    // target rect" ⇒ same raster. Bounds-checked eviction clears wholesale.
    private sealed record PasteKey(byte[] Pixels, long Epoch, int SrcW, int SrcH, int DstW, int DstH, bool Round, bool Mirror);
    private static readonly Dictionary<PasteKey, VideoFrame> _pasteCache = new();
+    private static readonly Dictionary<SceneElement, VideoFrame> _elementRasters = new(ReferenceEqualityComparer.Instance);
    private static readonly object _pasteGate = new();
-    private const int MaxPasteEntries = 48;
+    private const int MaxPasteEntries = 256;

    /// <summary>
    /// Composite <paramref name="scene"/> into the tier's output frame. Layer order (back →
@@ -294,14 +295,26 @@ public sealed class SceneCompositor
            // Rasterize in ELEMENT space on a TRANSPARENT base: the sampler writes 255
            // alpha only where it actually draws, so untouched margins (outside the
            // round clip, past the UniformToFill edges) stay alpha 0 and the paste
-            // reveals the master beneath instead of black.
-            var pixels = new byte[dw * dh * 4];
-            BlitContentRaw(pixels, dw, dh, 0, 0, ew, eh, frame, 1f, isRound, mirror);
-            raster = new VideoFrame(dw, dh, pixels);
+            // reveals the master beneath instead of black. The ARRAY is reused per
+            // element (see _elementRasters note) — new source generations re-raster
+            // in place, so the steady state allocates nothing.
            lock (_pasteGate)
            {
+                if (_elementRasters.TryGetValue(element, out var existing)
+                    && existing.Width == dw && existing.Height == dh)
+                    raster = existing;
+            }
+            if (raster == null) raster = new VideoFrame(dw, dh, new byte[dw * dh * 4]);
+            Array.Clear(raster.BgraPixels);
+            BlitContentRaw(raster.BgraPixels, dw, dh, 0, 0, ew, eh, frame, 1f, isRound, mirror);
+            lock (_pasteGate)
+            {
+                // Only the SOURCE frame's identity is a real key owner; drop any older
+                // entry that shared this element's reused raster so a stale key can never
+                // paste a mid-update array.
                if (_pasteCache.Count >= MaxPasteEntries) _pasteCache.Clear();
                _pasteCache[key!] = raster;
+                _elementRasters[element] = raster;
            }
        }

@@ -254,6 +254,69 @@ public class SceneCompositorTests
        Assert.Equal(out2.BgraPixels, out3.BgraPixels);
    }

+    /// <summary>take-16 regression gate — GC churn convicted mechanically: a ring-style
+    /// producer hands out a small set of rotating arrays with a NEW EPOCH every frame
+    /// (exactly what the capture/camera rings do). The compositor must re-rasterize INTO
+    /// its reused per-element raster, so steady-state allocation per render is near zero.
+    /// Before the fix every new cache key minted a fresh raster array (~625KB for the
+    /// webcam at 30fps ≈ 18MB/s) — gen2 pauses followed, and they ate camera frames in
+    /// the real app. The content assertion rides along: the pasted pixels must match the
+    /// CURRENT ring generation, never a stale one.</summary>
+    [Fact]
+    public void PasteCache_RingRecycledSources_ReRasterInPlace_WithoutAllocationChurn()
+    {
+        const int SrcW = 640, SrcH = 360;
+        var ring = new byte[4][];
+        for (var s = 0; s < ring.Length; s++)
+        {
+            ring[s] = new byte[SrcW * SrcH * 4];
+            for (var i = 0; i < ring[s].Length; i += 4)
+            {
+                ring[s][i] = 0; ring[s][i + 1] = 0;
+                ring[s][i + 2] = (byte)(40 + s * 50);
+                ring[s][i + 3] = 255;
+            }
+        }
+
+        var background = new Source { Type = SourceType.DisplayCapture, IsBackground = true, CaptureKey = "monitor:0" };
+        var webcam = new WebcamSceneConfig { X = 100, Y = 100, Width = 320, Height = 180 };
+        var scene = new Scene { Name = "Live" };
+        scene.Elements.Add(background);
+        scene.Elements.Add(webcam);
+        var options = new CompositorOptions
+        {
+            SourceRectX = 0, SourceRectY = 0, SourceRectWidth = 1280, SourceRectHeight = 720,
+            OutputWidth = 1280, OutputHeight = 720,
+        };
+
+        var bg = Solid(1280, 720, 0, 255, 0);
+        var compositor = new SceneCompositor();
+        long epoch = 0;
+        VideoFrame? FrameFor(SceneElement e) => e switch
+        {
+            WebcamSceneConfig => new VideoFrame(SrcW, SrcH, ring[epoch % ring.Length])
+                { IsOpaque = true, Epoch = ++epoch },
+            _ => bg,
+        };
+
+        var scratch = new byte[1280 * 720 * 4];
+
+        for (var i = 0; i < 8; i++) compositor.Render(scene, FrameFor, null, options, scratch: scratch);
+
+        var before = GC.GetAllocatedBytesForCurrentThread();
+        const int iterations = 200;
+        VideoFrame? last = null;
+        for (var i = 0; i < iterations; i++) last = compositor.Render(scene, FrameFor, null, options, scratch: scratch);
+
+        var perFrame = (GC.GetAllocatedBytesForCurrentThread() - before) / (double)iterations;
+        Assert.True(perFrame < 16 * 1024,
+            $"compositor allocated {perFrame:F0} bytes/frame under a ring producer — raster reuse regressed");
+
+        Assert.NotNull(last);
+        AssertColor(last!, 260, 190, (byte)(40 + 3 * 50), 0, 0);
+        AssertColor(last!, 5, 700, 0, 255, 0);
+    }
+
    [Fact]
    public void Render_VerticalTier_Outputs_1080x1920_From_The_Center_Crop()
    {