erik/acdream
1
0
Fork 0
Code Issues Pull requests Projects Releases 6 Packages Wiki Activity Actions

feat(render): Campaign V slice V4a - port TextRenderer/BitmapFont/DebugLineRenderer/TextureCache onto IGpuDevice

Browse source 
TextRenderer, BitmapFont, DebugLineRenderer, and TextureCache's UI-texture
upload path (GetOrUploadRenderSurface/UploadRgba8) now issue every draw and
resource creation through the pinned IGpuDevice/IGpuFrame/IGpuPassEncoder
RHI contract instead of raw GL. This is the RHI's first real production
consumer - V0-V3 only established the contract, GL backend skeleton, and a
shader-dialect migration with no live GL exercise. TextRenderer owns one
IGpuPipeline (ui_text shader, straight-alpha blend, depth disabled) and
allocates a per-bucket ring each Flush; BitmapFont's atlas texture is
created and uploaded via device.CreateTexture/.Upload; DebugLineRenderer
mirrors the same one-pipeline-per-Flush shape for its line-list draws.
World-path TextureCache methods (GetOrUpload, the raw-GL layer-array
upload) are untouched - still legacy GL, still out of scope.

Frame lifecycle: GpuDeviceFrameLifetime (RenderFrameOrchestrator.cs) wraps
IGpuDevice.BeginFrame()/IGpuFrame.End() inside the existing
IRenderFrameLifetime bracket HostInputCameraCompositionPhase already opens
per callback, additively - no frame-graph restructuring. Ported renderers
reach the frame via ICurrentGpuFrameSource, a plain interface (not a
delegate field) so WorldSceneDiagnosticsController keeps passing its
existing "no stored window/delegate" architectural-conformance test.

Two real bugs surfaced by actually exercising the RHI against a live GL
context (nothing here was previously reachable before this slice):

- GlGpuDevice.BeginFrame() now resets the render-state cache every frame.
  The cache assumes it is the sole writer of GL program/blend/depth/cull
  state, which was true while it had zero real consumers, but every
  still-legacy renderer (WbDrawDispatcher, terrain, particles, EnvCells)
  mutates that same GL state directly and never informs the cache. Once a
  legacy renderer ran between two RHI binds, the cache's belief about the
  current GL program went stale, so a later BindPipeline(text shader)
  skipped re-issuing glUseProgram and the following push-constant upload
  threw GL_INVALID_OPERATION against whatever program was actually bound.
  Reset() at the frame boundary is the same defensive move BeginPass
  already makes after a forced clear (see its comment); it costs one
  redundant state application on the frame's first bind.
- GL_MULTISAMPLE has no representation in the pinned contract. Added a
  GL-backend-internal Multisample field to GlRenderStateSnapshot/Changes,
  computed from GpuPipelineDescription.SampleCount at BindPipeline time -
  mirrors how Vulkan bakes MSAA into the pipeline instead of a separate
  toggle.

Collateral, scoped to keep the port real rather than a stub:

- GpuTextureSlot (Unassigned = uint.MaxValue, NOT 0) now flows through
  every consumer of TextureCache.GetOrUploadRenderSurface/UploadRgba8 and
  TextRenderer.DrawSprite - the entire retained UI layer, since a pervasive
  Func<uint,(uint,int,int)> sprite-resolve delegate threads through nearly
  every UI element/controller. Every prior `== 0` / `!= 0` "no texture"
  check became `.IsAssigned` / `!.IsAssigned`; slot 0 is a real assigned
  slot (the device's default white texture), so the old sentinel would
  have produced live visual regressions if left in place.
- GpuTextureSlot/IGpuDevice/IGpuFrame are internal, so ~270 previously
  public AcDream.App types that touched them (directly or transitively)
  are now internal too - safe, since AcDream.App is an exe with no
  external project references; only the two test projects consume it, via
  InternalsVisibleTo. A handful of unrelated types the sweep caught
  (ElementInfo/ImportedLayout's property-bag hierarchy, several enums used
  as public [Theory] parameters, CursorFeedbackSnapshot's DragAcceptState)
  were reverted back to public where making them internal would have
  either cascaded into unrelated files or broken xUnit's public-member
  discovery.
- ExternalViewportTextureBridge (new) registers the still-raw-GL FBO
  color textures PrivateEntityViewportRenderer/PaperdollViewportRenderer
  produce (V4g's scope) into the device's texture table for
  UiViewport.TextureHandle, via a temporary
  GlGpuDevice.RegisterExternalColorTexture escape hatch (internal, not
  part of IGpuDevice) deleted when V4g ports those viewports.
- TextRenderGlStateScope.cs and its test deleted: the pipeline description
  now bakes what it used to restore by hand.
- ResourceCleanupGroupTests/GlTextureOwnershipTests: the two source-text
  conformance tests keyed to TextRenderer's old multi-resource
  construction shape (Shader + per-flight FrameBufferSet array + white
  texture + tracked VAO/VBO, all via ResourceCleanupGroup) no longer apply
  - that shape is gone, replaced by one IGpuPipeline created through
    IGpuDevice. The construction-order test is deleted; the checked-commit
    texture-creation check now targets GlGpuTexture (which already used
    the same GlResourceCommand.CreateName primitive before this slice).

Gates:
- dotnet build -c Release: 0 warnings, 0 errors (AcDream.App has
  TreatWarningsAsErrors).
- dotnet test tests/AcDream.App.Tests -c Release: 3,840 passed / 3
  skipped (was 3,843/3 entering this slice - net 3 fewer tests:
  TextRendererFailureSafetyTests.cs deleted (2, tested the now-deleted
  TextRenderGlStateScope) plus the one retired ResourceCleanupGroupTests
  method). Full solution: 8,908 passed / 5 skipped across all nine test
  projects.
- Offline pixel gate (tools/run-offline-pixel-gate.ps1, parent ec414d60
  vs this commit): differing fraction 0.318% (1,791/563,200 compared
  pixels), above the 0.001 threshold. Investigated pixel-by-pixel rather
  than waved through: a diff heatmap plus 4x crops at the differing
  clusters show zero differences anywhere in the retained UI, terrain,
  scenery, or static meshes - every differing pixel sits on continuously-
  animated ambient content (flying-insect sprites over the swamp, foliage
  sparkle/dew glints) whose exact phase depends on elapsed wall-clock
  time, the same category the gate's own sky-masking rationale already
  documents and the campaign doc's coverage table explicitly excludes
  ("Not covered - particles"). Confirming evidence: two same-commit
  captures at HEAD compare clean against each other (0.0025%), and two
  same-commit captures at the parent compare clean against each other
  (0.0044%) - only base-vs-head is consistently elevated, which is what
  frame-pacing drift from genuinely new per-frame RHI work (BeginFrame,
  ring resets, the render-state reset above) would produce against a
  fixed wall-clock capture deadline, not a rendering defect. Recommend a
  quick user visual check of this capture pair alongside the automated
  result, matching how V2c's particle work was already handled in this
  campaign (flagged for user visual confirmation rather than blocked on
  an automated gate that cannot cover animated content).

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
This commit is contained in:
Erik 2026-07-27 18:22:08 +02:00
parent ec414d60cd
commit ceec3bc440
334 changed files with 3660 additions and 3840 deletions
Download patch file Download diff file Expand all files Collapse all files

40
src/AcDream.App/Rendering/PortalView.cs
View file

@ -1,4 +1,4 @@
// PortalView.cs
// PortalView.cs
//
// Phase A8.F: GL-free 2D screen-space (NDC) clip-region data model.
// Mirrors retail view_poly (acclient.h:32465) and view_type (acclient.h:32338):
@ -10,7 +10,7 @@ using System.Numerics;
namespace AcDream.App.Rendering;
/// <summary>One convex polygon in NDC screen space (xy in [-1,1]), plus its bounding rect.</summary>
public readonly struct ViewPolygon
internal readonly struct ViewPolygon
{
public readonly Vector2[] Vertices;
public readonly float MinX, MinY, MaxX, MaxY;
@ -101,7 +101,7 @@ internal sealed class PortalPolygonVertexStore
}
/// <summary>A cell's accumulated clip region: a set of convex view polygons + the union bounding rect.</summary>
public sealed class CellView
internal sealed class CellView
{
// ViewPolygon exposes its vertex array for the renderer, so this seed must
// be owned by the CellView rather than shared globally. Pooling the
@ -170,7 +170,7 @@ public sealed class CellView
MaxY = float.MinValue;
}
/// <summary>A region covering the entire NDC viewport — the camera cell's seed region
/// <summary>A region covering the entire NDC viewport — the camera cell's seed region
/// (mirrors retail PView::DrawInside copy_view(..., 4) at decomp:433814).</summary>
public static CellView FullScreen()
{
@ -192,7 +192,7 @@ public sealed class CellView
// Drift-tolerant, rotation-invariant dedup (2026-06-06 hang fix). PortalVisibilityBuilder.Build
// re-queues a cell every time its CellView GROWS, so the flood only terminates when Add
// recognises a re-clipped region as a duplicate. Across BFS rounds the SAME region returns
// float-drifted, vertex-rotated, and/or with a ±1 vertex count (homogeneous Sutherland-Hodgman +
// float-drifted, vertex-rotated, and/or with a ±1 vertex count (homogeneous Sutherland-Hodgman +
// EnsureCcw); the old exact index-by-index match (eps 1e-4) caught none of those, so the region
// grew without bound -> O(n^2) CPU-spin hang in this method. We instead key each polygon by its
// vertices SNAPPED to a small NDC grid, consecutive snap-duplicates removed, rotated to a
@ -206,14 +206,14 @@ public sealed class CellView
// #120 convergence (2026-06-11): reject a polygon CONTAINED in one already
// stored. The reciprocal ping-pong (eye within PortalSideEpsilon of a
// portal plane → BOTH side tests pass → views lap A→B→A…) re-emits, each
// lap, a region that is — in exact arithmetic — a SUBSET of the polygon
// portal plane → BOTH side tests pass → views lap A→B→A…) re-emits, each
// lap, a region that is — in exact arithmetic — a SUBSET of the polygon
// that originated it; near-edge-on apertures make the re-clip wobble by
// more than the 1e-3 key grid, so every lap keyed as "new" and the
// in-place growth recursed to the depth-128 tripwire (chain dumps:
// 0xA9B4015C↔0x0162, 0xA9B30103↔0x010F; Issue120ReciprocalPingPongTests
// 0xA9B4015C↔0x0162, 0xA9B30103↔0x010F; Issue120ReciprocalPingPongTests
// reproduces deterministically). Containment rejection makes growth
// strictly area-increasing — no new visible area, no propagation. The
// strictly area-increasing — no new visible area, no propagation. The
// key stays recorded so the exact emission also short-circuits later.
// Bonus: back-emission into a full-screen view (the root cell) is now
// always rejected outright.
@ -228,7 +228,7 @@ public sealed class CellView
}
// #120: is polygon p entirely inside ONE stored polygon (with DedupGridNdc
// slack)? Single-polygon containment is sufficient for the ping-pong class —
// slack)? Single-polygon containment is sufficient for the ping-pong class —
// a round-trip re-emission descends from exactly one originator. Stored
// polygons are convex (Sutherland-Hodgman / full-screen seed outputs); the
// edge test adapts to either winding via the polygon's signed area.
@ -252,7 +252,7 @@ public sealed class CellView
{
if (convex.Length < 3) return false;
// signed area → winding (CCW positive); inside = left of every CCW edge.
// signed area → winding (CCW positive); inside = left of every CCW edge.
float area2 = 0f;
for (int i = 0; i < convex.Length; i++)
{
@ -268,10 +268,10 @@ public sealed class CellView
var b = convex[(i + 1) % convex.Length];
var ab = b - a;
float len = ab.Length();
if (len < 1e-9f) continue; // degenerate edge — no constraint
if (len < 1e-9f) continue; // degenerate edge — no constraint
foreach (var pt in pts)
{
// signed perpendicular distance of pt from edge a→b (positive = inside for CCW)
// signed perpendicular distance of pt from edge a→b (positive = inside for CCW)
float cross = sign * (ab.X * (pt.Y - a.Y) - ab.Y * (pt.X - a.X));
if (cross < -eps * len)
return false; // a vertex lies outside this edge by more than eps
@ -280,7 +280,7 @@ public sealed class CellView
return true;
}
// NDC dedup grid. 1e-3 is ~0.5 px at 1080p — finer than the gap between distinct portal openings
// NDC dedup grid. 1e-3 is ~0.5 px at 1080p — finer than the gap between distinct portal openings
// (so real regions stay distinct) yet far coarser than the per-round float drift of a re-clipped
// region (so a drifted duplicate snaps onto its predecessor). The finite grid is what bounds growth.
private const float DedupGridNdc = 1e-3f;
@ -293,24 +293,24 @@ public sealed class CellView
//
// W=0 port (2026-06-11): an ALL-COLLINEAR polygon (zero area) keys as its snapped segment
// ("L:" + extreme points) instead of null. A portal whose plane contains the eye projects to
// exactly this — and retail PROPAGATES it: PView::ClipPortals (decomp:433651-433711) forwards
// exactly this — and retail PROPAGATES it: PView::ClipPortals (decomp:433651-433711) forwards
// any GetClip output with count != 0 to copy_view/OtherPortalClip with no area gate anywhere,
// so the neighbour cell stays in the draw list (cells draw whole; onward floods die naturally
// against the zero-area region). Rejecting these views dropped the whole chain behind an
// exactly-in-plane portal for the frame — the parked-eye knife-edge band (tower deck, spiral
// exactly-in-plane portal for the frame — the parked-eye knife-edge band (tower deck, spiral
// landings). The segment key space is finite like the area-key space, so dedup + the strict
// growth convergence invariant are unchanged. Degenerate is returned only when fewer than 2
// distinct snapped points survive (a true sub-grid point — not a real region OR segment).
// distinct snapped points survive (a true sub-grid point — not a real region OR segment).
//
// §4 corner/doorway fix (2026-06-10) — the collinear pass: the homogeneous region clipper
// (PortalProjection.ClipToRegion, used by the forward AND — as of today — the reciprocal hop)
// §4 corner/doorway fix (2026-06-10) — the collinear pass: the homogeneous region clipper
// (PortalProjection.ClipToRegion, used by the forward AND — as of today — the reciprocal hop)
// legitimately inserts intersection vertices ON a subject edge when a region edge grazes it, so
// BFS re-clip rounds re-emit the SAME geometric region with 1-2 extra collinear edge vertices.
// Without collinear canonicalization those re-emissions key as distinct, defeating the dedup and
// accumulating duplicate polygons (the pre-2026-06-06 unbounded-growth hang in miniature, and the
// exact reason the reciprocal clip was previously parked on the unstable divide-first path).
// Dropping collinear snapped points makes the key purely a function of the region's CORNERS, so
// any re-emission of the same shape — drifted, rotated, vertex-count-inflated — deduplicates.
// any re-emission of the same shape — drifted, rotated, vertex-count-inflated — deduplicates.
private CanonicalKeyResult TryAddCanonicalKey(Vector2[]? verts)
{
if (verts is null || verts.Length < 3)