Commit Graph
6274 Commits
Author SHA1 Message Date
Khawar Zamman Wani 830cb65d48 fix(segment): place cancellation tests after items 2026-10-28 19:53:59 +05:30
Khawar Zamman Wani 5a86877b68 fix(segment): honor cancellation on short BK cuts 2026-10-28 19:35:31 +05:30
Khawar Zamman Wani a79057f6e4 perf(segment): solve GrabCut and the band re-cut with Boykov-Kolmogorov
GrabCut's colour-model grids have few terminal links and short paths, where
BK beats the push/relabel solver the default dispatch picks for graphs of
4096+ nodes. Quick Selection keeps push/relabel.

24 MP synthetic, release, median of 3: Select Subject 480 -> 207 ms,
Object Selection 337 -> 140 ms, Quick Selection unchanged. IoU unchanged.
2026-10-28 17:01:21 +05:30
aiongg 2b1c6b93c3 Merge main into split-filter-adjust-crop-fx-builders
crop_ui.rs moved to tools/; the PR's builders replace main's Clippy extractions of the same code; #6723's Color Lookup save carried into the moved encoder.
2026-10-28 04:22:40 +00:00
aiongg b4e69af424 Merge main into perf/9328-fallback-reason 2026-10-28 03:00:01 +00:00
Dylan M. Taylor d5be9b8aef Merge pull request #9770 from dylanmtaylor/sonar/enforce-clippy-standards
Enforce Rust coding standards via Clippy
2026-10-27 22:55:30 -04:00
aiongg 594a28b335 Render path: a raised GPU Memory ends GPU Memory Full; a flipped view keeps its own slot (#9328)
A measured fit that fits ends the document's over-budget reason, so the status bar clears as
soon as GPU Memory is raised. A flipped or pixel-aspect-corrected view records its reason in
its own slot: unflipped, the document's own reason (Advanced Blending...) is back rather than
a false "back on the GPU".
2026-10-28 02:19:44 +00:00
Dylan M. Taylor ab9b27b80d Restore rationale comments dropped by the Clippy refactors 2026-10-28 02:13:42 +00:00
Dylan M. Taylor 163caeb86b Make the PSD size-limit test fail when a size-based PSB promotion happens 2026-10-28 02:13:42 +00:00
Dylan M. Taylor 63eb3eabe4 Drop the unused band_rows export and its dead_code allow from io 2026-10-28 02:13:42 +00:00
Dylan M. Taylor c726a14f0f Restore the Animated Zoom hold gating in canvas navigation 2026-10-28 01:49:51 +00:00
Dylan M. Taylor 1a3415608b Restore the PSD frame stroke import after the io refactor 2026-10-28 01:49:51 +00:00
Dylan M. Taylor 5ad52e6b85 Remove duplicated view-command and type-target code in ui-egui 2026-10-28 01:06:43 +00:00
Dylan M. Taylor f2ac9d28b7 Refactor ui-egui layer and channel panels for strict Clippy checks 2026-10-28 01:06:43 +00:00
Dylan M. Taylor 267fb55e0f Refactor ui-egui canvas and options bar for strict Clippy checks 2026-10-28 01:06:42 +00:00
Dylan M. Taylor de9636f56e Refactor ui-egui dialogs for strict Clippy checks 2026-10-28 01:06:42 +00:00
Dylan M. Taylor 1a6d8ec9e3 Refactor ui-egui control dispatch for strict Clippy checks 2026-10-28 01:06:42 +00:00
Dylan M. Taylor a703ea2781 Restore wasm builds of ui-egui after the Clippy refactor 2026-10-28 01:06:42 +00:00
Dylan M. Taylor 9800f56dfb Refactor text for strict Clippy checks 2026-10-27 23:47:07 +00:00
Dylan M. Taylor 6b397ff7c0 Refactor tablet for strict Clippy checks 2026-10-27 23:47:06 +00:00
Dylan M. Taylor 52800a7b3d Refactor raw for strict Clippy checks 2026-10-27 23:47:06 +00:00
Dylan M. Taylor ba25f585bd Refactor psd for strict Clippy checks 2026-10-27 23:47:06 +00:00
Dylan M. Taylor 2b0b11848c Refactor paint for strict Clippy checks 2026-10-27 23:47:06 +00:00
Dylan M. Taylor 2a3135d796 Refactor io for strict Clippy checks 2026-10-27 23:47:06 +00:00
Dylan M. Taylor 6ebc0180da Refactor gpu for strict Clippy checks 2026-10-27 23:47:06 +00:00
Dylan M. Taylor 2ed75b4ec3 Refactor geom for strict Clippy checks 2026-10-27 23:47:06 +00:00
Dylan M. Taylor 169bbdfa22 Refactor format for strict Clippy checks 2026-10-27 23:47:06 +00:00
Dylan M. Taylor af45d71447 Refactor engine for strict Clippy checks 2026-10-27 23:47:06 +00:00
Dylan M. Taylor f0ffcef216 Refactor compose for strict Clippy checks 2026-10-27 23:47:06 +00:00
Dylan M. Taylor 9a348da8bb Refactor codecs for strict Clippy checks 2026-10-27 23:47:06 +00:00
Dylan M. Taylor ddaf299335 Refactor automation for strict Clippy checks 2026-10-27 23:47:06 +00:00
Dylan M. Taylor 8bc939cf3b Refactor algo for strict Clippy checks 2026-10-27 23:47:06 +00:00
Dylan M. Taylor 32f884d75c Refactor UI tools for strict Clippy checks 2026-10-27 23:47:05 +00:00
Dylan M. Taylor dccf6182eb Refactor UI support for strict Clippy checks 2026-10-27 23:47:05 +00:00
Dylan M. Taylor 79e9fa6f7c Refactor UI panels for strict Clippy checks 2026-10-27 23:47:05 +00:00
Dylan M. Taylor 211d4824c8 Refactor UI menus for strict Clippy checks 2026-10-27 23:47:05 +00:00
Dylan M. Taylor 97fc920872 Refactor UI input for strict Clippy checks 2026-10-27 23:47:05 +00:00
Dylan M. Taylor bba45115fa Refactor UI dialogs for strict Clippy checks 2026-10-27 23:47:05 +00:00
Dylan M. Taylor 9366cd58ec Refactor UI canvas for strict Clippy checks 2026-10-27 23:47:05 +00:00
Dylan M. Taylor ef065ee66e Refactor UI application for strict Clippy checks 2026-10-27 23:47:05 +00:00
Brennan StehlingandClaude Opus 5.5 285944de8d Perf: P79's row name leaves out the document size, which --quick changes
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-27 14:38:45 -07:00
Brennan StehlingandClaude Opus 5.5 e43762ac0d Perf: rename P79 to the reduced composite it times (#7494)
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-27 14:38:45 -07:00
Brennan StehlingandClaude Opus 5.5 72b53ffebe Perf: P78 times a hide on every run, showing the layer again untimed first (#7494)
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-27 14:38:45 -07:00
Brennan Stehling 6beffc0fb9 Perf: P78 forces the GPU budget to 512 MB, so hiding the top adjustment layer takes the CPU fallback (#7494) 2026-10-27 14:38:45 -07:00
Brennan StehlingandClaude Opus 5.5 8cecf418e5 Perf: cliff rows for the CPU compositor, Advanced Blending, adjustment toggles and a huge document (#7494)
Reported only (enforce = false), so the off-happy-path cliffs have a "before":
- perf_scenarios_cpu: the layered rows on the forced CPU compositor (P68-P74, #9330)
- P75 opacity with Advanced Blending on another layer (#6689), P76 adjustment toggle (#6535)
- P77 hiding an adjustment among many full-canvas layers (#6533), added after P60/P61 run
- P78 zoomed-out reduced refresh of a 30000x30000 painted document (#1015, #9330)

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-27 14:38:45 -07:00
Brennan StehlingandClaude Opus 5.5 1382a240a6 Render path: an adjustment preview on a software adapter records its reason (#9328)
A software adapter still runs the GPU canvas, so an adjustment dialog that
previews through the CPU proxy is its own cause, as a document's reason is.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-27 12:20:25 -07:00
David MannandClaude Opus 5.5 5ccb6fb723 Merge main
Conflicts with main's batch #9880:
- psd file.rs / layer.rs: #5287 (PSB by encoded size, RLE tables widened on
  promotion) against #6583's lazy reader. Main's new widening and length code
  read `.data`, which #6583 leaves empty while the bytes stay in the source;
  it now goes through `encoded()` / `encoded_len()` (also the auto-merged
  lines in `rle_widening`, `encoded_len` and `LayerRecord::write`), and the
  lazy reader sets `write_as_psb: false`. New test
  `a_lazy_psd_promotes_to_the_same_psb_as_an_eager_one`: a lazily read PSD
  writes the same PSB bytes as an eager one; it fails with a `.data` read.
- direct_select.rs / vector_ui.rs: #9321's Type Path and #6353's hidden
  Work Path both kept.
- io tests main.rs: both modules. scorecard.md regenerated (12 of 181);
  ui-parity.md and gaps.md follow it.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-27 19:49:42 +01:00
David Mann 8bc71f924b Merge main
# Conflicts:
#	crates/format/src/store.rs
2026-10-27 19:10:28 +01:00
aiongg 55effab1aa Merge #9689: CMYK: the test reference keeps a fully covered type pixel's ink exact (#9688) 2026-10-27 17:44:31 +00:00
aiongg 0bc6abc783 Merge #9680: Plug-ins: refuse a code section with no functions but many bytes instead of panicking (#9679) 2026-10-27 17:44:30 +00:00