* GPU: free effect working textures after each call instead of at frame end
A GPU frame keeps its encoder open while the CPU renderer runs layers with GPU
effects (Gpu::effects -> run_chain). Each call's Enc handed its textures back
to the shared pool, where they waited for every open encoder, so all effect
intermediates of a frame stayed allocated until the frame ended (17.7 GB by
layer 17 of a heavy 4K comp, then OOM). Effect calls now borrow a pool owner
(Enc::lent) that is given back once the call's work is submitted or
discarded; its kept textures are trimmed to the pool budget, so the rest are
freed after each call.
* GPU: one pool budget for idle effect calls, trimmed at the end of a render
Review of #1523: each idle lent owner kept up to POOL_BUDGET of textures of
its own, outside the shared accounting, and there were as many idle owners as
effect calls ever ran at once (16 frame threads: ~12 GB after a 4K render).
Giving a lent owner back now moves its kept textures to the shared free list,
so one POOL_BUDGET bounds what the free list and every finished effect call
keep. They skip Pool::release's wait for open encoders: only the lent Enc
recorded work on them, and that work is submitted or discarded. Each lend is
a fresh owner id, so a lease that outlives its call (an image handed to
another encoder) is released through Pool::release; this does not rely on
the result being read back.
Accelerator::render_end (called when an export ends) drops the kept textures.
* GPU: evict the pool's oldest free textures, guard escaped leases
Review of #1523 (follow-up to 0a884fb6):
- Pool::make_free rejected a texture when the free list was full and never
evicted, so once sizes that no longer occur (per-frame blur bounds, padded
layer sizes) filled the budget, every later free was a drop and every take
an allocation. It now evicts the least recently freed textures until the
new one fits.
- A lease read by another Enc (dispatch, copy, readback, display) or handed
to a cache is marked escaped and always released through the guarded
Pool::release, so a lent call can no longer reuse a texture that a still
open encoder reads.
- Docs: POOL_BUDGET caps the free list only; textures waiting for open
encoders and those of live Encs are outside it.
- export_async calls render_end from a drop guard, so the pool is also
trimmed when an export panics or its future is dropped.
- Tests use a private context with its own pool budget instead of a global
override; new tests for stale sizes and an image read by an open encoder.
* GPU: evict only stale pool sizes, count evicted textures until the GPU frees them
Evicting the least recently freed texture for every newcomer thrashed when a frame
frees more than the budget in a repeating order (blur/glow intermediates): each free
evicted the size the next call takes first, so misses came back to the old level and
evict + reallocate raised the peak.
The free list now evicts only sizes nobody asked for in the last 2 output frames
(Accelerator::frame_end, called by the export per frame); otherwise the incoming
texture is dropped and the kept set keeps hitting. Evicted bytes count against the
budget until the work submitted so far is done, so evictions run at most one budget
ahead of the GPU.
* GPU: lower the shared pool budget to 256 MB
On 4K RTX 3090 runs 256 MB matches the old code's peak reserved on d_blurglow
(9204 MB) and is 256-512 MB above it on c_effects (6592 vs 6080 MB), while
staying faster.
* GPU: end frames on the multi-frame render paths, route recording through Enc::record
Only the export pipeline called frame_end, so off export (preview, remote
frames) the pool's clock stood still, no size went stale and a full free
list dropped every newcomer. passes::comp_frame (exports), remote frames
and the preview workers now end each frame; test without pool_frame_end.
Docs: the frame clock is coarse with many frames in flight; dying only
throttles eviction, approximately; no cache path escapes images. Enc::record
is the single way to get the encoder outside context.rs and marks the
images it names (Enc::uses).
* GPU: wait for the empty submission while evicted pool textures still count
stale_sizes_do_not_starve_a_steady_size failed with 0 hits and 20 misses on
hardware adapters: pool_frame_end submitted an empty batch and then only did a
non-blocking Poll, and the queue's work-done callback did not run before the
next frame, so Pool::dying never drained and make_free stopped evicting stale
sizes.
While dying is non-zero, pool_frame_end now waits on that empty submission for
up to the readback timeout (the pool lock is already released). A wait timeout
just moves on and does not retire the context or its readbacks; other poll
errors still retire them. Without dying textures it keeps the cheap Poll.
Enc::crop records through Enc::record(&[src, &dst]) so both pooled images are
registered like on every other recording path (review).
* GPU: do not repeat the pool frame-end wait until evicted textures drain
A wait that frees none of the evicted bytes (timeout, or a WebGPU poll that
does not block) marks the pool stalled; later frames only poll until dying
drops. Test covers the no-repeat and partial-progress cases; dying decrement
saturates.
* Remote frames: end a pending frame that a new render replaces
A render arriving while a multi-pass frame waits for readbacks dropped that frame without
`frame_end`, so the GPU pool's clock skipped it. Errors return before `frame_begin`.
* Render: end the frame even when a comp frame panics
`passes::comp_frame` called `frame_end` only after a normal return, so a frame that
panicked (or whose future was dropped between passes) never aged the GPU pool.
* GPU pool: log the reuse rate at render end; test reuse across effect calls and frames
Log cumulative hits/misses (and reuse %) once per render before trimming. Tests
cover reuse within an open frame across effect calls, and across repeated frames.
---------
Co-authored-by: Matej Kozlovský <MatejKozlovsky@users.noreply.github.com>
Co-authored-by: bflatastic <bombayapirat@gmail.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: FredApps <153173854+FredApps@users.noreply.github.com>
The default studio environment's level-preserving mean of 1 acted like a
100% white Ambient light: with only the Environment light on, a white floor
rendered at full white (95% of the frame saturated) while its shadow core
went to pitch black — a fog over the scene, nothing like After Effects'
default. Three changes bring the look in line, CPU and GPU alike:
- The default environment is a dim studio now: a solid-angle-weighted mean
of 0.3 (and a stronger key gradient), so a white floor reads a mid gray
and the light's colour x intensity still scales the level.
- The shadow occludes only the image's directional (key) fraction of the
light — |sum of dir x radiance| / sum of radiance, computed beside the
dominant direction and carried in the scene — so the fill keeps shining
inside the shadow: an area-light shadow, partial like After Effects'
(a lone bright spot on black still shadows almost fully; a uniform image
not at all).
- The environment shadow map's penumbra has a floor — eight output pixels,
or four map texels where those are coarser — because the map covers the
whole scene bounding sphere rather than the view, and its texels would
otherwise stair-step the edge (all the more in a draft render's flat 512
budget). Shadow Diffusion widens it further, as before.
Tests: the key fraction tracks directionality; the default keeps its dim
level, its direction and its key; the default and a bright key both shadow
the floor (the default partial, the key deep) with Cast Shadows off lit
everywhere; and the full look — a mid-gray lit floor (no fog), a partial
core, and a soft edge ramp in quality and draft renders alike.
The Casts Shadows option Environment lights got is only half a fix if
nothing shades for it: an image lights the scene from every direction,
and what occludes that light is the image itself — its dominant bright
region, the sun in a sky map.
- dominant_direction finds it: the radiance-weighted mean texel
direction (solid-angle weighted, rotation applied). A uniform image
cancels to zero — a flat default environment shines from everywhere
at once and casts no shadow, which a uniform environment test pins.
- The shadow is one orthographic map over the scene from that
direction, after the lights' maps, like a parallel light's: PCF
widened by Shadow Diffusion, darkened by Shadow Darkness, cleared
when nothing casts, the image is directionless, or the texel budget
drops it.
- It darkens the environment's own diffuse and specular light on
Accepts Shadows surfaces and counts in Cast Shadows Only. CPU and
GPU (WGSL) do the same; Globals carries the environment's sun
direction and shadow settings (40 → 44 floats, preflight updated).
A bright-disc environment turned to shine from behind the camera now
shadows a sphere onto the floor behind it, Cast Shadows off lights it
evenly again, and the GPU scene's sky gradient takes a shadow-casting
Environment light. Also: the renderer-switch toasts read "...renderer,
which draws 3D models", the dialog's Casts Shadows checkbox gets an
auto id (dialog.light.casts) the tests click, and the docs' Environment
notes cover the shadow.
The Advanced 3D renderer folded Accepts Shadows (Off/On/Only) into a
bool, so a shadow catcher set to Only drew as an ordinary lit surface.
The material now carries shadows_only. Such a fragment draws black at
the strongest shadow a light casts on it, as Classic 3D does, and it
goes in the blended pass so what is behind it shows through. The CPU
shader and advanced3d.wgsl get the same branch.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
On Adreno (Windows ARM64) a failed vkAllocateMemory surfaces as an internal
ERROR_INITIALIZATION_FAILED, not out-of-memory. Gpu::within_memory only scoped
OutOfMemory, and wgpu's DeviceExt::create_buffer_init maps the new buffer at
creation and panics ("Failed to get mapped range") when the allocation failed,
taking down a frame thread. Buffers with initial contents are now created unmapped
and filled through queue.write_buffer (context::buffer_init), so a failed
allocation is only an invalid buffer and an error for the caller's scope, and
within_memory scopes Internal errors too, discarding the frame for the CPU path
as it does for out-of-memory.
The device was then lost, and egui-wgpu panicked on the main thread staging the
next frame's index buffer on the dead device. Once the host reports the
presentation device lost, the app draws nothing more (no mesh to stage), auto-saves
unsaved changes, keeps the crash-recovery marker so the next launch offers them,
and closes its window.
Tests: a device with a small buffer limit gets an over-limit and an unaligned
buffer_init without a panic and sees the error in its scope; the UI harness
reports a device loss and checks the next frame has no shapes, closes the window
and wrote the auto-save.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01QiAJeFttZp83Fh7YuzAR5f
FreeBSD's CI VM (Mesa llvmpipe over GL) failed 22 effectcraft-gpu tests:
- Advanced 3D's camera-depth target was R32Float, which GLES can't render to without
EXT_color_buffer_float: create_render_pipeline failed validation (release builds only log
that and draw nothing). The target is now R32Uint holding the f32 bits (renderable on every
backend, lossless), and GpuContext checks the adapter's format features (render attachment,
blendable colour) and independent blending up front; when they're missing the render
pipelines are never built and Advanced 3D renders on the CPU.
- Signed `%` in the shaders: naga emits GLSL `%`, undefined for negative operands, which broke
the VR effects' horizontal wrap and Cell Pattern's tiling. A shared `imod` never takes `%` of
a negative number.
- asin/acos: Mesa's are coarse approximations (~1e-4 rad), visible in every lens/sphere/cylinder
mapping. `asin_p`/`acos_p` (Cephes' asinf polynomial) are f32-accurate on every backend.
- "Could not lock adapter context": the GL backend has one context lock per adapter with a
timeout, so tests driving devices concurrently panic. Every gpu test helper that hands out a
device holds one process-wide lock for the calling test; ui_web_job_gpu's two device tests
share a lock too.
- On GL, up to 1 % of pixels may flip across a decision boundary (division through a
reciprocal, unfused fma); the GPU-vs-CPU comparisons allow that there only.
Also adds the FreeBSD workflow (from the FreeBSD CI PR) so this branch is checked there.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
- GPU in browser job workers: each job worker opens its own WebGPU device
(deferred readbacks); the Render Queue export and the analyses' frame loops
are futures (render::passes, offload::run_request_async) whose frames
render in passes, awaiting the device between them. GPU particles and
Advanced 3D read back under keys too. workers.jobs reports passes and
readbacks; the smoke test measures them.
- Layer buffers in the browser's OPFS disk cache: frame workers back their
layer caches with a PrefetchStore; the page plans what to fetch before a
request from recent misses (LayerPrefetch over the shared DiskIndex).
- Templates embed footage (default on, size-capped with a warning) and
creating from a template extracts it next to the new project / into
browser storage.
- ICC LUT profiles (lut8/lut16/lutAtoB/lutBtoA, from the public ICC spec) in
Simulate Output > My Custom RGB, baked into a 3D LUT for the viewer.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
- GPU in the browser: each frame worker opens its own WebGPU device (Gpu::request_deferred) and
renders the GPU compositor and GPU effects with deferred readbacks (effectcraft-gpu deferred):
readbacks keyed by content start with map_async and miss, the pass returns placeholders, the
layer cache's gate drops inserts, and FrameServer::resume renders again once the device
delivered. Backend Auto times whole frames per comp in the workers. Frames with effects go to
GPU workers instead of the page. CPU fallback without WebGPU in workers (?nogpuworkers).
- Fix: a WGSL constant (3.40282347e38) that Chrome rejects invalidated every GPU kernel in Chrome
(the page's compositor too); a test checks every float literal fits f32.
- Disk cache in the Origin Private File System: workers write sealed frame entries with sync
access handles (temp file + move), the page keeps the DiskIndex (LRU under the settings' limit,
at most half the quota) and reads hits instead of rendering; survives reloads; blue cache bar.
- Storage manager: Settings ▸ Disk ▸ Browser Storage (usage/quota, persistent storage, Clear
buttons, automation ids) and storage.info / storage.persist / storage.clear (StorageHost).
- Docs: web.md (GPU workers, deferred passes, disk cache, storage manager, threading design:
one engine instance per worker), architecture, preferences, parity (WEB-1), roadmap.
- Tests: deferred passes vs CPU on a real device, placeholder gating, frame server passes and
Auto, GPU worker routing + disk cache across a reload, storage commands and widgets, store
usage by kind; smoke test: GPU effect frames in workers with pixel parity, disk hits after a
reload, storage manager.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
render: three_d::adv::SplitRun (Renderer::split_adv_run) hands draw_run's split path to an
accelerator as data (main scene, special layers far to near, 3D mattes drawn solo);
Renderer::sky_draw / three_d::SkyDraw for Environment Light Background layers. gpu: the walk
draws skies (adv_sky kernel), resolves each special layer, hides it behind the nearer main
scene (occlude kernel) and composites it through the 2D path without readbacks.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Advanced 3D runs render entirely on the GPU: every motion-blur sub-sample
rasterised, resolved and averaged, the depth-based iris depth of field
(Classic 3D bokeh spans, highlight boost, progressive levels), the sRGB
encode and the composite over the GPU canvas (Accelerator::render_3d,
Renderer::prepare_adv_run, adv3d.wgsl). Wireframe outlines draw on the GPU.
Fixes NaN environment lookups for straight up/down normals on Metal.
bench --gpu gains an Advanced 3D comp (--adv3d).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Advanced 3D comps render their 3D runs as one scene of world-space triangles: model and
primitive layers (meshes, skinning, animation clips by layer time), extruded text/shape
layers, and textured cards for every other layer with their Material Options. Shading is
physically based in linear light: Cook-Torrance GGX + Lambert for parallel/spot/point lights
(AE falloff and cone), ambient lights, image-based light from an Environment light's layer
(equirectangular; cosine irradiance map + roughness mip chain + split-sum BRDF fit), and
PCF shadow maps (spot/parallel/point cube faces). 2x2 SSAA, depth of field as a gather blur
by circle of confusion, back to the working encoding.
The software rasteriser (depth buffer, top-left rule, perspective-correct, opaque visibility
then sorted transparent blending) is the reference and CI path; effectcraft-gpu implements
Accelerator::raster_3d with a wgpu render pipeline (advanced3d.wgsl) and matches it within
tolerance. Environment layers are hidden from both renderers.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>