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

feat(render): Campaign V slice V6l commit 1 - particles draw on Vulkan

Browse source 
Contract amendment 1 of three, and V4e's content behind it. Plan section 5.5.16
recorded that both particle pipelines draw with per-instance VERTEX attributes
and that the pinned contract could express instanced DRAWING but not instanced
vertex INPUT: one stride, no divisor, one buffer at VertexInputRate.VERTEX. That
is what stopped V4e. This takes the reviewed option (i) - a second vertex
binding with a per-instance rate.

The amendment. GpuVertexLayout grows a per-binding notion (binding index,
stride, input rate) and GpuVertexAttribute names the binding it is fed from,
defaulting to 0; IGpuPassEncoder.BindVertexBuffer takes a binding index. Every
layout written before this slice keeps its exact meaning through
GpuVertexLayout.Interleaved, which is one vertex-rate binding 0 - and
GpuContractTests asserts that as a requirement rather than trusting it. Both
backends carry the rate natively and at no cost: VK_VERTEX_INPUT_RATE_INSTANCE
on the pipeline, glVertexAttribDivisor recorded once into the pipeline's VAO
where it survives every later attribute rebind.

GpuVertexFormat.UInt1 comes with it, and is necessary to it: particle.vert
declares `layout(location = 6) in uint aTextureIndex` and the amendment's whole
premise is that no shader is edited. Same kind-distinction UByte4UInt was added
for at V4d - GL needs glVertexAttribIPointer, Vulkan needs R32_UINT, and the
float path would reinterpret the value's bits rather than approximate them.

Options (ii) and (iii) were rejected on the record: all ten storage bindings are
spoken for and reusing binding 0 would have the GL particle draw clobber
WbDrawDispatcher's instance array mid-frame (section 5.5.8's hazard in its GL
form); CPU-expanding instances is 5x billboard bandwidth and does not scale to
mesh particles at all.

The arm. ParticleRenderer.Rhi.cs is a SECOND arm per section 5.5.6, not a
replacement - every GL statement in the sibling file is the one it always
issued. Five pipelines replace the imperative glBlendFunc switch (two billboard
blends, three mesh blends) because core Vulkan 1.3 does not make blend dynamic.
The per-flight VAO/VBO pool disappears because every ring allocation inside a
frame is already distinct memory that lives until the frame retires. The
binding-9 table is not bound at all - the device owns the table and the encoder
binds set 2. The pass is BORROWED from IWorldPassScope. Depth tests but does not
write, compare is Less and alpha-to-coverage is off, which is the ambient GL
state particles have always drawn under rather than a choice. Everything above
the submission seam - emitter iteration, retail distance ordering, the
deferred-alpha handoff, billboard axis construction, blend resolution - is the
same CPU code on both arms.

The first Vulkan particle frame threw rather than drew, which is the second
defect of the compiles-clean class this slice found by running:
TextureCache.AcquireParticleTexture is bindless-only, so the standalone particle
texture cache did not exist on a backend without GL. It exists on both arms now.
Everything about it that matters - sharing equivalent surfaces between emitter
owners, the bounded unowned LRU, retirement behind the frame-flight fence - is
already backend-neutral; only how one entry is created and destroyed differs,
which is what IStandaloneBindlessTextureBackend is for. The RHI arm creates the
image through IGpuDevice.CreateTexture with a real sampler and releases the
table slot before the image, which is the GL arm's order and for the same
reason. The composite cache stays GL-only: it serves entity appearance, not
particles.

The durability fix V6k earned. That slice found the sky declaring a 32-byte
stride against a 36-byte AcDream.Core.Terrain.Vertex - the record carries a
TerrainLayer no sky attribute names - and noted that every .Rhi.cs arm restates
a CPU record's footprint from memory while only sky had a test.
RhiVertexLayoutStrideTests is that test for the rest: world mesh, terrain, sky,
retained-UI sprite, debug line, and both particle bindings, each asserted
against the record or the producer's own float count, plus two sweeps over all
seven for attributes that reach past their stride or name an undeclared binding.
Four private layouts became internal to be assertable; nothing else about them
moved.

Gates. Release build green. App tests 4,121/3 skips (4,109 baseline plus three
contract tests and nine layout tests); complete Release suite 9,184/5. Strict GL
offline pixel gate against 08ffe141: 3.20e-05, 18 differing pixels of 563,200,
inside the documented 9-31 band. GL connected -Runs 3: 3/3 RENDERED on the
desktop witness and 3/3 on the client capture. One offline Vulkan run with
VK_LAYER_KHRONOS_validation proven inserted by the loader: zero validation
errors, zero warnings, a captured world frame that still draws terrain,
blending, roads, water, statics, scenery, sky and the complete retained UI.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
This commit is contained in:
Erik 2026-07-28 17:20:59 +02:00
parent 08ffe141a0
commit b1ad1d481b
23 changed files with 1360 additions and 64 deletions
Download patch file Download diff file Expand all files Collapse all files

110
src/AcDream.App/Rendering/TextureCache.cs
View file

@ -187,6 +187,22 @@ public sealed unsafe class TextureCache
_compositeTextures = composite;
_particleTextures = particles;
}
else
{
// Campaign V slice V6l: particles draw on both arms, so the standalone
// particle cache exists on both. Everything about it that matters —
// sharing equivalent surfaces between emitter owners, the bounded
// unowned LRU, and retirement behind the frame-flight fence — is
// backend-neutral; only how one entry is created and destroyed
// differs, which is what the backend interface is for. The COMPOSITE
// cache stays GL-only: it serves entity appearance, not particles,
// and its port is not this slice's.
_particleTextures = new StandaloneBindlessTextureCache(
new ParticleRhiTextureBackend(this),
retirementQueue,
budgets.StandaloneUnownedBytes,
budgets.StandaloneUnownedEntries);
}
}
/// <summary>
@ -546,6 +562,16 @@ public sealed unsafe class TextureCache
surfaceId,
origTextureOverride: null,
paletteOverride: null);
// Campaign V slice V6l: the RHI arm has no GL name to intern a bindless
// handle from, so the image is created through IGpuDevice and paired
// with a real sampler object. The decode above is the same one the GL
// arm uses, so the pixels are identical; the shader still samples layer
// zero of a one-layer array, which is what a Texture2D registered into
// the table is on Vulkan (VulkanTextureFormatMapping.SampledViewTypeOf).
if (_gl is null)
return AcquireParticleTextureRhi(textures, ownerId, surfaceId, decoded);
uint name = UploadRgba8AsLayer1Array(decoded);
ulong handle = 0;
try
@ -604,6 +630,54 @@ public sealed unsafe class TextureCache
}
}
/// <summary>
/// Campaign V slice V6l: one particle surface as a device texture-table
/// slot, owned by the same emitter-scoped cache the GL arm uses.
///
/// <para>Linear/clamped is the filtering the GL arm's own one-layer array
/// upload sets on itself, and a particle sheet's UVs never leave [0,1] —
/// the quad's own texcoords are the unit square — so the wrap mode is not a
/// visible choice, it is just the safe one.</para>
/// </summary>
private GpuTextureSlot AcquireParticleTextureRhi(
StandaloneBindlessTextureCache textures,
uint ownerId,
uint surfaceId,
DecodedTexture decoded)
{
IGpuTexture texture = _device.CreateTexture(new GpuTextureDescription(
$"particle-surface-0x{surfaceId:X8}",
GpuTextureKind.Texture2D,
GpuTextureFormat.Rgba8Unorm,
Width: decoded.Width,
Height: decoded.Height,
LayerCount: 1,
MipLevelCount: 1));
try
{
texture.Upload(0, 0, decoded.Rgba8);
uint accountingName = UploadAccountingName(texture);
TrackUploadedTexture(accountingName, decoded.Width, decoded.Height);
IGpuSampler sampler = _device.CreateSampler(GpuSamplerDescription.WorldClamp);
GpuTextureSlot slot = _device.RegisterTexture(texture, sampler);
textures.AddAndAcquire(ownerId, new StandaloneBindlessTextureResource
{
SurfaceId = surfaceId,
Name = accountingName,
Texture = texture,
Slot = slot,
Bytes = checked((long)decoded.Width * decoded.Height * 4L),
});
return slot;
}
catch
{
texture.Dispose();
throw;
}
}
internal void ReleaseParticleTextureOwner(int emitterHandle)
{
if (emitterHandle <= 0 || _particleTextures is null)
@ -768,11 +842,14 @@ public sealed unsafe class TextureCache
return _compositeTextures!;
}
private StandaloneBindlessTextureCache EnsureParticleTexturesAvailable()
{
EnsureBindlessAvailable();
return _particleTextures!;
}
/// <summary>
/// Campaign V slice V6l: no longer gated on bindless. The particle cache is
/// constructed on both arms, so the only failure left is a cache that was
/// never built at all.
/// </summary>
private StandaloneBindlessTextureCache EnsureParticleTexturesAvailable() =>
_particleTextures ?? throw new InvalidOperationException(
"This TextureCache owns no standalone particle texture cache.");
private sealed class ParticleTextureBackend(TextureCache owner)
: IStandaloneBindlessTextureBackend
@ -792,6 +869,29 @@ public sealed unsafe class TextureCache
=> owner.DeleteUploadedTexture(resource.Name);
}
/// <summary>
/// Campaign V slice V6l: the same ownership boundary on a backend with no
/// bindless handles. The table slot is released first and the image second,
/// which is the same order the GL arm uses and for the same reason — a
/// submitted-but-unretired frame may still sample the slot, and
/// <see cref="IGpuDevice.ReleaseTextureSlot"/> is what defers its reuse.
/// </summary>
private sealed class ParticleRhiTextureBackend(TextureCache owner)
: IStandaloneBindlessTextureBackend
{
public void MakeNonResident(StandaloneBindlessTextureResource resource)
{
if (resource.Slot.IsAssigned)
owner._device.ReleaseTextureSlot(resource.Slot);
}
public void Delete(StandaloneBindlessTextureResource resource)
{
resource.Texture?.Dispose();
owner.UntrackUploadedTexture(resource.Name);
}
}
/// <summary>
/// Advances bounded composite-cache maintenance once per render frame.
/// Logical owner release is immediate; at most one over-budget layer and