The inter-pass and acquire/present barriers in the frame loop set both
stage masks to ALL_COMMANDS, and the inter-pass dependency used a
queue-wide VkMemoryBarrier — fully serialising against every pipeline
stage and flushing all caches every frame, when all the next pass needs
is the swapchain image the previous one wrote.
Replace the inter-pass global VkMemoryBarrier with an image memory
barrier scoped to the swapchain image's single colour subresource (as
the intra-pass UI barrier already does), and derive the barrier stage
masks per pass: RenderPass::SwapchainStage() is overridden by UIRenderer
(COMPUTE_SHADER) and RTPass (RAY_TRACING_SHADER), so a compute->compute
edge only serialises COMPUTE while an RT pass pulls in RAY_TRACING — the
acquire/present frame-edge masks use the real union of the frame's
passes (SwapchainStageUnion). The base default and the empty-passes
fallback are the conservative COMPUTE | RAY_TRACING | TRANSFER union, so
a polymorphic or un-overridden pass can only be over- not
under-synchronised.
Adds SwapchainBarrierScope (pure CPU) pinning the per-pass derivation,
the union narrowing, and the inter-pass barrier scope; FrameLoopSync
already drives the real GPU frame loop with validation enabled.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
vertexBuffer/indexBuffer/aabbBuffer were HOST_VISIBLE (system RAM), so the
BLAS build, every refit, and any hit-shader geometry fetch read them over
PCIe. The usage flags (SHADER_DEVICE_ADDRESS, plus STORAGE on the index
buffer) exist precisely to expose geometry for hit-shader fetch, the dominant
RT shading pattern.
Add VulkanBuffer::UploadDeviceLocal, which places a CPU-written/GPU-read
buffer in device-local memory and picks the upload mechanism at runtime via
the #89 strategy (Device::PreferDirectDeviceWrite):
- ReBAR/UMA: allocate HOST_VISIBLE|DEVICE_LOCAL (DEVICE_LOCAL preferred),
map transiently, memcpy, flush-if-non-coherent. No staging buffer.
- no/small BAR: allocate pure DEVICE_LOCAL + TRANSFER_DST, fill a transient
HOST_VISIBLE staging buffer, vkCmdCopyBuffer, then DeferredClear the
staging buffer onto the fence-keyed deletion queue (#101/#102) so it
outlives the copy submit.
The geometry buffers become non-mapped so the destination is free to be
device-local-only. Same-size re-uploads reuse the allocation, so the device
address stays stable across an in-place AS UPDATE refit. The compressed Build
path now allocates vertex/index as pure DEVICE_LOCAL — the GPU decompressor
fills them directly, no host write.
Tests: BLASBuildOptions asserts device-local placement on the triangle,
procedural, and (budget-forced) staged paths, and exercises the staged copy +
deferred-deletion + in-place refit under GPU-assisted validation with zero
validation errors.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The shader binding table is written once at pipeline creation and read by
the GPU on every vkCmdTraceRaysKHR thereafter, yet it was allocated plain
HOST_VISIBLE (system RAM) — so every trace dispatch read raygen/miss/hit
records over PCIe.
Route the SBT allocation through Device::PreferDirectDeviceWrite (#89):
when a DEVICE_LOCAL|HOST_VISIBLE type exists, prefer DEVICE_LOCAL on top
of the required HOST_VISIBLE so the buffer lands in mappable device
memory (GPU reads hit local VRAM; the one-time memcpy goes write-combined
over PCIe). When no combined type exists, the preferred hint is dropped
and GetMemoryType (#59) falls back to plain HOST_VISIBLE — current
behaviour. The SBT is tiny so a small BAR window's budget is never the
constraint.
The buffer stays mapped and the existing FlushDevice gates on the chosen
memory type's coherency flags (#60), so a direct-write type lacking
HOST_COHERENT is still flushed correctly.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The compressed Mesh::Build / ImageVulkan::Update paths kept their host-visible
`compressedStaging` (holding the GDeflate streams) alive for the whole life of
the mesh/image, pinning host-visible memory long after the single GPU decompress
that reads it has retired.
Release it via VulkanBuffer::DeferredClear() right after recording the decompress.
The recorded vkCmdDecompressMemoryEXT still references compressedStaging.address,
so it must outlive the submit — exactly what the fence-keyed deletion queue
(#101/#102) guarantees: it retires the allocation only after framesInFlight frames
have elapsed, by which point the submit's fence has cleared. Between Builds/Updates
the handle is null; the next call's Resize re-creates it.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
A hardware acceleration-structure UPDATE cannot change topology, so the
index buffer is immutable on the Refit UPDATE path. Re-uploading it was a
wasted memcpy + FlushDevice + barrier of unchanged data every refit —
genuinely per-frame for deforming meshes.
Skip the index memcpy and its FlushDevice on the UPDATE branch; the
RecordBLASBuild call still reads the stable, unchanged indexBuffer.address.
The `indicies` span is now read only for its count. Document the API
nuance: a bitwise-different but topologically-equivalent index array is
ignored on refit, consistent with "a refit may only move vertex positions."
The full-rebuild fallback path still re-uploads both buffers.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Since #40 dropped the per-frame vkQueueWaitIdle, frames are pipelined:
a resource the CPU is done with may still be read by the GPU for up to
numFrames-1 more frames. VulkanBuffer::Resize's destroy-and-recreate
path was therefore a live use-after-free (#63), no longer masked by the
wait-idle.
Device gains a monotonic, frame-counter-keyed deletion queue:
EnqueueDeletion tags {buffer, memory} with the current frameCounter;
ReclaimDeletions (called per frame after the fence wait) frees entries
once framesInFlight frames have elapsed; DrainDeletions frees everything
after a wait-idle. VulkanBuffer::DeferredClear hands handles to the queue
and nulls the handle, and Resize uses it instead of immediate Clear().
Window::Render sets framesInFlight at init, reclaims after the per-image
fence wait, bumps Device::frameCounter once per frame, and drains on the
resize / OUT_OF_DATE wait-idle paths. The destructor keeps immediate
Clear() (callers destroying mid-flight remain responsible, unchanged).
Adds the DeferredDeletion test: drives the retire timing on a real
headless device with real buffers, stepping Device::frameCounter to pin
the exact reclaim frame, asserting validation stays silent.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The mip-generation upload path issued N+1 layout-transition barriers per
chain. The final blit's destination level is never read again, so its
dedicated TRANSFER_DST -> TRANSFER_SRC barrier was redundant: the level was
transitioned to SRC only to keep the final SRC -> consumer transition uniform.
Skip that last per-level barrier and fold the final level into the closing
transition, which now batches two VkImageMemoryBarrier entries
([0, mipLevels-1) from TRANSFER_SRC, the last level from TRANSFER_DST) into a
single vkCmdPipelineBarrier. One fewer barrier call per chain; the N-1
interleaved per-level barriers stay one-at-a-time since each is read by the
next blit. Applied to both the host-upload and GPU-decompress Update paths.
Adds the MipChainBarrierBatch regression test driving the pure barrier
builder (no GPU device needed).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
WriteBufferDescriptor / WriteSampledImageDescriptor / RegisterSampler /
WriteSwapchainDescriptors each FlushDevice()'d the entire multi-KB per-frame
descriptor heap after writing one few-byte descriptor. Add a ranged
VulkanBuffer::FlushDevice(offset, bytes) that flushes only the touched bytes,
rounding the range outward to nonCoherentAtomSize as vkFlushMappedMemoryRanges
requires (the WHOLE_SIZE path sidestepped that). The coherent-memory gate from
issue #60 is preserved — coherent memory still skips the flush entirely.
The descriptor writers now self-flush their written range, so the redundant
whole-heap flushes in UIRenderer::Initialize and the resize callback are gone.
Rounding lives in AlignMappedFlushRange (pure math) and is unit-tested without
a device in the new VulkanBufferRangedFlush test; the change was also verified
end-to-end by running HelloUI on a real GPU with validation layers (font-atlas
glyphs, sampler and storage-image descriptors all flush correctly, no VUIDs).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
vkCreateComputePipelines and vkCreateRayTracingPipelinesKHR were passed
VK_NULL_HANDLE, so the driver couldn't reuse compiled shader binaries
across the pipelines built at startup (4 UI shaders + user/RT pipelines)
or across runs.
Add one process-wide Device::pipelineCache. LoadPipelineCache (called
from Initialize) seeds it from pipeline_cache.bin when the file's
VkPipelineCacheHeaderVersionOne header matches this device's
vendorID / deviceID / pipelineCacheUUID; a stale or foreign blob is
discarded so the driver never rejects it. SavePipelineCache is registered
with std::atexit, serialising the cache at process shutdown without any
explicit teardown hook. Both create sites now pass the cache; a null
handle stays valid, so no create-site null check is needed.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Resize previously always destroyed + reallocated. It now reuses the
existing allocation in place when the new request still fits the created
capacity and the immutable-at-create properties match: usage flags are
fixed at create, and the chosen memory type must still satisfy the
required property flags. preferredPropertyFlags is a best-effort perf
hint and is intentionally not part of the guard. On reuse the buffer
handle, device address and mapped pointer are preserved; only `size`
shrinks to the new logical extent.
Tracks capacity and the created usage flags on VulkanBufferBase, set at
Create and carried through the move constructor.
Adds VulkanBufferResizeReuse, a device-free regression test that drives
the reuse guard directly and asserts the in-place path issues no Vulkan
call.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Add Device::PreferDirectDeviceWrite(size) so the memory-placement cluster
(#65/#72/#73/#75) no longer re-derives where a CPU-written, GPU-read buffer
should live. It picks the strategy at runtime from Device::memoryProperties:
- No DEVICE_LOCAL|HOST_VISIBLE type (no resizable BAR) -> false (must stage).
- ReBAR/UMA (BAR heap >= 90% of largest DEVICE_LOCAL heap) -> true: map and
write directly; a staging copy would be pure overhead.
- Small BAR window (BAR heap << VRAM) -> true only for buffers within a
per-window budget (1/8 of the window); large buffers stage so they don't
exhaust the window (#58).
Complements GetMemoryType (#59): #59 picks the memory *type*, this picks the
*strategy*. Upload-only — readback (#74) wants HOST_CACHED and stays separate.
A VK_EXT_memory_budget-driven remaining-space check is noted as a follow-up.
Tested by tests/UploadStrategy (pure CPU over synthetic ReBAR / UMA /
small-window / no-BAR layouts, no GPU device needed), mirroring
MemoryTypeFallback.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
FlushDevice/FlushHost called vkFlushMappedMemoryRanges /
vkInvalidateMappedMemoryRanges unconditionally. Record the chosen
memory type's propertyFlags at Create time (via the index GetMemoryType
returns, not the requested flags) and early-return when HOST_COHERENT.
The cmd overload still always emits its pipeline barrier — only the
redundant host-side flush/invalidate is gated.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Add a fifth UI compute path — DispatchFused — alongside the four
per-element Dispatch* calls. It composites quads → circles → images →
text (canonical back-to-front order) in ONE dispatch that loads the
destination image once and stores once, eliminating the per-pass
load+store and the inter-pass VkMemoryBarriers a run of consecutive
Dispatch* calls would pay. The win materialises at >= 2 non-empty
categories; an absent category is a zero-trip, push-constant-uniform
loop (~free).
The new uber-kernel (shaders/ui-fused.comp.glsl) reuses each category's
exact cooperative shared-memory tile-cull + per-pixel accumulate, so a
fused category is pixel-identical to its standalone pass. Categories run
sequentially per thread and share one set of shared-memory scratch, so
the VGPR/shared high-water mark stays at the per-category level, not the
sum.
Additive and non-breaking: the per-element Dispatch* calls and the
frozen 48-byte UIDispatchHeader are untouched. DispatchFused gets its own
128-byte push-constant layout (UIFusedHeader: four uvec4 headers + four
per-category clip vec4s). To interleave a custom ui.Dispatch() between
standard passes, bracket it with two DispatchFused calls — the dispatch
boundary is the explicit, app-declared flush point.
WebGPU (Vulkan-second): DispatchFused falls back to the per-element
Dispatch* calls in canonical order — one texture/sampler per dispatch
there can't feed the bindless image phase — so the result matches; only
the load/store fusion is Vulkan-only.
HelloUI now composites its background quads, mouse circle, and label
text in a single DispatchFused. Verified rendering identical on a live
Vulkan/Wayland session.
Resolves#47
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Font::GetLineWidth called stbtt_GetCodepointHMetrics for every glyph on
every invocation. InputField_HitTestCursor rescans the whole field once
per character, making caret hit-testing O(n²) in HMetrics lookups.
Memoise advances per Font via Font::AdvanceUnits: ASCII in a flat array,
the rest in a map. Advances are cached in unscaled font *units* and
rescaled per call — Glyph::advance in the FontAtlas is baked at kBaseSize
and is wrong at any other size, so it cannot be reused. The per-glyph
(int) truncation in GetLineWidth is preserved exactly.
Adds the FontAdvanceCache CPU-only regression test covering size
independence of the cached unit advance, the per-glyph-truncated rescale
pipeline, cache-hit determinism, width scaling with size, and the
non-ASCII map path.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
uiResolveScreenPixel ran four float compares against hdr.clipRectPx for
every pixel of every UI dispatch, even though the clip rect is the
full-surface default {0,0,1e9,1e9} in the overwhelmingly common case.
Reserve the high bit of the header flags word (UI_FLAG_CLIP) and have
FillHeader set it only when the clip rect is narrower than the surface
(extracted into the unit-testable UIRenderer::ClipFlags). The shader
gates the compares on the bit; the branch is push-constant-uniform so it
never diverges. When the rect covers the surface the skipped compares
could never reject an in-surface pixel, so output is pixel-identical.
Mirrored on the WebGPU/WGSL path (dom-webgpu.js) for parity. Adds the
UIClipFlag CPU test covering the gate decision.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
InputField_HitTestCursor mapped a click x to a cursor byte offset by
calling Font::GetLineWidth on every prefix value[0..i] for i=1..size.
Each call re-walks from byte 0, so the hit test was O(n^2) glyph-metric
lookups (each an uncached stbtt cmap binary search). It also iterated
raw byte boundaries, so on multibyte UTF-8 input it could return an
offset splitting a codepoint — an invalid position for the byte-based
cursorPos.
Add Font::NearestCursorByte: one left-to-right cumulative-advance pass
(O(n) metrics, scale computed once) over codepoint boundaries, returning
the byte offset of the boundary nearest the target. Cumulative advance is
monotonic, so the scan stops past the closest boundary. The hit test now
delegates to it.
Adds tests/InputFieldHitTest, a pure-CPU regression test (no Vulkan
device/window): a px sweep across ASCII and multibyte strings compared
against an independent brute-force nearest-boundary oracle, plus the
left/right/empty/end-of-text edge cases and a codepoint-boundary invariant.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
GetMemoryType was a bare first-superset match that threw whenever no
memory type satisfied the requested property flags. Callers combining a
mandatory flag with a perf-only one (HOST_VISIBLE | DEVICE_LOCAL for the
descriptor heaps) would then fail allocation on any device without a
host-visible device-local heap (no resizable BAR).
It now takes a `preferred` mask distinct from `required`: a type
satisfying both is chosen first, falling back to a required-only match
when the preference is unavailable, and throwing only when even
`required` cannot be met. VulkanBuffer::Create/Resize gain an optional
trailing `preferredPropertyFlags`, and the descriptor heaps now treat
DEVICE_LOCAL as a preference on top of mandatory HOST_VISIBLE.
This does not touch any flush/barrier behaviour — coherency is not
assumed anywhere here.
Adds tests/MemoryTypeFallback, a pure-CPU test that installs synthetic
memory layouts and exercises selection, preference, fallback, and the
unsatisfiable-required throw.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
ImageVulkan hardcoded VK_PIPELINE_STAGE_RAY_TRACING_SHADER_BIT_KHR as the
consuming stage in every layout transition. The UI font atlas is sampled by
a *compute* shader, so the upload barrier's dst scope never covered the real
consumer — a correctness gap, not just lost overlap.
Parameterize the consuming stage per image: store it once at Create
(defaulting to RT, preserving existing RT/example callers) and reuse it in
every Update / UpdateRegion / compressed-Update transition. FontAtlas passes
VK_PIPELINE_STAGE_COMPUTE_SHADER_BIT.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Per glyph, ShapeText did cache_.contains(key) (in Ensure) then
cache_.find(key) (in Lookup) — two independent hashes + node-chases of
the same key. Add EnsureAndGet returning const Glyph* with a single
find, insert-on-miss, returning &it->second. It returns nullptr only on
ShelfPlace failure, preserving the existing `if (g==nullptr) continue;`
behavior. Ensure now delegates to it.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
ShapeText re-decoded UTF-8, re-looked-up every glyph, and recomputed
layout every frame for static strings (DrawText / EmitCenteredLabel run
inside per-frame onBuild). DrawText and EmitCenteredLabel also did a
second full pass over the glyphs to apply alignment.
Memoize the shaped run, keyed by (Font*, pxSize, color, utf8), as an
origin-relative vector<GlyphItem>. A cache hit skips decode/lookup/layout
and just translates the cached run into the output buffer by
(x + alignShift, baselineY). The alignment shift (Center = -advance/2,
Right = -advance) is folded into ShapeText's single pass, dropping the
second loop in both callers. TextAlign moves to the :UI partition so
ShapeText can take it.
Pure CPU memoization — the glyph buffer is fully rewritten and re-flushed
every frame, so a hit is byte-equivalent. A miss still calls Ensure so new
glyphs rasterise and dirty the atlas; hits don't (atlas entries are never
evicted). InvalidateFont drops a font's runs before it is destroyed (the
cache, like FontAtlas's glyph cache, keys on the Font pointer). A soft cap
bounds memory against pathological per-frame-unique text.
Adds the ShapeTextCache test (15 checks): hit == miss byte-for-byte, hits
don't touch the atlas, translate/alignment/truncation/InvalidateFont.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
FontAtlas::Update re-uploaded the entire 1024x1024 R8 atlas (1 MiB)
whenever a single glyph was rasterized, so warm-up / language-switch /
size-churn cost roughly one full-atlas copy per frame that added a
codepoint.
Accumulate a dirty bounding box (FontAtlas::DirtyRect) in MarkDirty as
glyphs are placed, and copy only that sub-rect. On Vulkan this is a new
ImageVulkan::UpdateRegion: the staging buffer keeps the full atlas row
stride, so bufferRowLength stays the atlas width and bufferOffset points
at the rect origin while imageOffset/imageExtent carve out the rect.
Layout transitions stay whole-image (cheap); only the copy extent
shrinks. WebGPU passes the rect straight to wgpuWriteAtlasRegion, which
already took x/y/w/h.
The freshly-zeroed atlas marks its whole extent dirty in Initialize so
the first Update clears the image once; thereafter every untouched texel
stays the zero it was uploaded as, keeping the image byte-identical to
the staging copy.
Tested: new FontAtlasDirtyRect unit test covers the accumulation /
clamp / lockstep-with-`dirty` math (no GPU needed); HelloUI renders text
correctly on both Vulkan (NVIDIA, validation-clean) and WebGPU.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The renderer was effectively single-buffered despite allocating
triple-buffered infrastructure: Render() ended with an unconditional
vkQueueWaitIdle, and a single (presentComplete, renderComplete)
semaphore pair with zero fences was shared for the Window's lifetime.
Rework the pacing model so up to numFrames frames overlap:
- Per-swapchain-image VkFence, signaled by the submit and waited+reset
before that image's command buffer / descriptor-heap slot is
re-recorded. Keyed by acquired image index (drawCmdBuffers, the heap
slots, and the swapchain images are all image-indexed — see
WriteSwapchainDescriptors). Created signaled so first use passes.
- Per-image render-finished (present) semaphore: the presentation engine
holds it until the image is re-acquired, so per-image is the only safe
key (per-CPU-frame trips VUID-vkQueueSubmit-pSignalSemaphores-00067).
- Per-CPU-frame acquire semaphores (image index unknown until acquire
returns), sized numFrames+1: in-flight depth is bounded by the
per-image fences, so numFrames+1 distinct acquire semaphores guarantee
the reused one has no pending op (VUID-vkAcquireNextImageKHR-01779).
- Drop the steady-state vkQueueWaitIdle; keep it on resize / OUT_OF_DATE
/ teardown.
Add tests/FrameLoopSync: drives the real frame loop against a live
Wayland compositor for 60 frames (>> in-flight slots) and asserts the
CPU frame counter advanced, the swapchain rotated across multiple
images, and the validation layer stayed silent — the load-bearing check,
since the old single-pair design only avoided being an active race
because the wait-idle masked it.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
CreateSwapchain unconditionally used VK_PRESENT_MODE_FIFO_KHR, but still
ran two vkGetPhysicalDeviceSurfacePresentModesKHR calls plus a heap-allocated
std::vector to enumerate present modes that were never consulted. Delete
them and assign the constant directly.
compositeAlpha is a surface property that does not change across swapchain
recreation, so select it once at construction instead of re-deriving it
(with another heap-allocated vector) on every resize configure. The
vkQueueWaitIdle before recreation is left intact as required.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Render() rebuilt two VkBindHeapInfoEXT structs every frame and
recomputed reservedRangeOffset via a runtime subtract + bitmask
against Device::descriptorHeapProperties. The heap address/size per
slot are stable for a given heap; only currentBuffer varies.
Precompute numFrames resource/sampler bind structs and cache them,
keyed on the descriptorHeap pointer. The cache rebuilds whenever the
heap is (re)assigned and otherwise just indexes by currentBuffer.
Pointer-keyed invalidation catches any heap reassignment — the same
point a setter hook would fire — and is independent of onResize, as
the heaps never resize today (per the issue's correctness caveat).
Resolves#42
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Window::Render() rebuilt cmdBufInfo, the subresource range, both
VkImageMemoryBarrier structs, and presentInfo on the stack every frame
even though only the barriers' image (and the acquire barrier's
oldLayout) ever change. Hoist them to members initialised once in the
constructor; Render() now patches only the two varying fields.
presentInfo.pImageIndices/pSwapchains point at the currentBuffer and
swapChain members, so they track value changes (including swapchain
recreation) automatically. The render-complete wait semaphore is fixed
for the window's lifetime, so the per-frame `renderComplete != NULL`
check — set once, never cleared — was dead and is removed; the
semaphore is wired into presentInfo in the constructor instead.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Replace the std::vector<nanoseconds> frameTimes plus
erase(begin()) with a fixed-size std::array ring buffer (head +
count). The previous code memmoved ~99 elements left every frame once
the window filled. LogTiming computes order-independent sum/avg/min/max,
so the ring's write order is irrelevant to the reported stats.
Behind CRAFTER_TIMING only; shipping builds are unaffected.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Add zero-copy procedural BLAS overloads that take the AABB build input
straight from an existing device buffer instead of memcpy-ing a host span
through Mesh::aabbBuffer — the input-source companion to #36's refit work.
A GPU compute producer (e.g. a GPU-resident particle system) can now feed a
moving procedural BLAS that builds/refits each frame with no host round-trip.
Vulkan: new BuildProcedural(VkDeviceAddress, count, ...) and
RefitProcedural(VkDeviceAddress, count, ...) feed the device address directly
into VkAccelerationStructureGeometryAabbsDataKHR; the host-span path is
refactored to share RecordProceduralBuildFromAddress and is otherwise
unchanged.
WebGPU: device-buffer BuildProcedural/RefitProcedural copy the boxes GPU->GPU
into the mesh heap (new wgpuRegisterMeshBLASDeviceAabbs / wgpuRefitMeshBLAS-
DeviceAabbs bridge fns) and wrap them in a single root leaf bounded by a
caller-supplied worldBounds — no wasm round-trip, blasAddr stable across refit.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
