e2e/CI: dual-engine suites, per-engine screenshots, SwiftShader retired, prod web suite, CI-coverage gate

Squash of experiment/ff-big-modules vs main.

Big-module routing removed: native-EH shrank kicad_editor below
SpiderMonkey's x86-64 code budget (runs 29355049705/29356152413 green on
stock Firefox), so BIG_MODULE_SPECS routing and the baseline-only-JIT
crutch are gone — kicad-firefox and kicad-chromium both run the full
suite, with the module compiled the way real users' browsers compile it.

Per-engine screenshots end to end: stableShot/shotPath write
test-results/<engine>/<name>.png; baselines move to
baseline-screenshots/{chromium,firefox}/ and the whole tools/screenshots
pipeline (compare/promote/manifest/spec-map/changelog/Discord) keys on
<engine>/<name>. Previously Firefox and Chromium renders of one spec
overwrote each other and Firefox renders were never actually gated.
Seeded from CI run 29421380806 (92 new firefox baselines, +24 chromium
web-suite shots); manifest generated from the baseline tree.

One merged playwright.config.ts (kicad/asyncify/coroutine/perf as
projects); ~25 dead npm scripts dropped. The web suite is gated in CI for
the first time ever (4 rotted specs fixed, 5 broken lib-bridge specs
triaged as fixme in docs/features/web-e2e-rot/); cheap lint step after
npm ci; last 26 blind-sleep violations fixed.

SwiftShader retired: CI Chromium renders WebGL on ANGLE → Mesa llvmpipe
(--use-gl=angle --use-angle=gl --ignore-gpu-blocklist; the blocklist flag
is mandatory — llvmpipe is blocklisted and WebGL is silently unavailable
without it) in BOTH configs. Under WORKERS=4 congestion SwiftShader
transiently failed the first post-board-load draw and the recovery
cascade ended in a silent permanent Cairo fallback — that engine flip was
the "~1.2% changedRatio both directions" occ-export baseline flake.
Validated 160/160 across two 80-repeat rigs; full analysis in
docs/features/wx-parity-bugs/occ-export-context-eviction.md. Chromium
baselines shift slightly on llvmpipe — promote once from the first green
run. Deflakes the new coverage exposed: presence baselines settle before
capture; presence fixtures declare current file formats; perf gets its
own outputDir so CI evidence survives; occ-export settles the board paint
before the export dialog; menu-item waits (waitForRenderedByLabel before
clickMenuItem) in 4 specs + the TESTING.md rule.

Web suite runs the PROD build, in parallel: webServer becomes backend
`start` + the standalone's e2e:preview (build-preview.mjs: link-wasm →
stash the public/wasm symlink aside during vite build, build-demo.mjs's
move — then vite preview as the persistent server). The wasm middleware
serves /wasm/* in preview and emits COOP/COEP/CORP itself (a pthread
worker script's own response must carry COEP or Chrome kills it with
ERR_BLOCKED_BY_RESPONSE). VITE_* flags bake at build time;
VITE_ALLOW_USER_OVERRIDE joins turbo globalEnv. fullyParallel + default
workers: 5.2m → 1.4m. Determinism fixes the parallel run exposed:
shared-page specs become serial groups; locks.spec grabs alice's exact
item via the new kicadCollabTestSelectByUuid hook (cross-tab "first
footprint" order is not a ysync invariant); quit specs poll page.url()
(quit supersedes its own navigation — NS_BINDING_ABORTED on Firefox).
Suite: 51 passed / 12 skipped / 0 failed in 1.6m.

CI-coverage gate (lint:ci-coverage): every tests/**/*.spec.ts must be
reachable from the npm scripts the workflows invoke — scraped from
.github/workflows/, resolved through package.json, coverage asked from
playwright --list itself. Rules: uncovered-spec + orphan-project (with a
documented LOCAL_ONLY_PROJECTS allowlist). Gating next to
lint:determinism; 138 spec files / 13 projects accounted for.

Product fixes kept from the investigations (reachable on real GPUs too):
wx 7799fd1be5 — paint flags clear before dispatch + Invalidate always
propagates; kicad 3dcfea5e45 — SwiftShader pass-boundary flush +
per-instance font texture + first-frame GL-error drain (GAL recovery
recovers instead of falling back to Cairo) + the user-facing eeschema
switch navigates again under __EMSCRIPTEN__ (project-sync's
FaceRegistered gate had rerouted it into the hidden sync player; caught
by the newly-gated web suite).

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_018eUxiPApHgGiu9NFyQfhAq

tests/e2e/aui-resize.spec.ts

@@ -21,7 +21,7 @@
 import { test, expect, waitForWxApp, MAIN_CANVAS } from './utils/fixtures';
 import {
   findRenderedByType,
-  findAuiPaneContent, stableShot } from './utils/element-tracker';
+  findAuiPaneContent, stableShot, shotPath } from './utils/element-tracker';
 import type { Page } from '@playwright/test';
 
 const APP = '/standalone/aui/aui_test.html';
@@ -242,11 +242,11 @@ test.describe('wxAuiManager dock-sash resize UX (pcbjam#20)', () => {
     // intermediate frame while the button is still held.
     await page.mouse.move(sash.x - 90, sash.y, { steps: 10 });
     await page.waitForTimeout(70); // eslint-disable-line -- documented interaction dwell: let the mid-drag frame paint before capturing it for the pixel-diff
-    const midShot = await page.screenshot({ path: 'test-results/aui-resize-03-mid-drag.png', scale: 'css' });
+    const midShot = await page.screenshot({ path: shotPath(page, 'aui-resize-03-mid-drag.png'), scale: 'css' });
 
     await page.mouse.up();
     await page.waitForTimeout(200); // eslint-disable-line -- documented interaction dwell: let the release settle before capturing the after frame for the pixel-diff
-    const afterShot = await page.screenshot({ path: 'test-results/aui-resize-04-after.png', scale: 'css' });
+    const afterShot = await page.screenshot({ path: shotPath(page, 'aui-resize-04-after.png'), scale: 'css' });
 
     const midFrac = await stippleFraction(page, midShot, region);
     const afterFrac = await stippleFraction(page, afterShot, region);

tests/e2e/contextmenu.spec.ts

@@ -1,7 +1,7 @@
 // Context-menu (DoPopupMenu) e2e — right-click pops a DOM popup, items route
 // wxEVT_MENU back to the handler, and outside-click/Escape cancels.
 import { test, expect, tryLoadApp, getCanvasBox } from './utils/fixtures';
-import { findRenderedByType, clickMenuItem } from './utils/element-tracker';
+import { findRenderedByType, clickMenuItem, shotPath } from './utils/element-tracker';
 
 async function rightClickCanvasCentre(page: import('@playwright/test').Page) {
   const box = await getCanvasBox(page);
@@ -45,7 +45,7 @@ test.describe('DOM-port context menu', () => {
     expect(paste, 'Paste should be present').toBeTruthy();
     expect(paste!.enabled, 'Paste should be disabled').toBe(false);
 
-    await page.screenshot({ path: 'test-results/contextmenu-01-open.png', fullPage: true });
+    await page.screenshot({ path: shotPath(page, 'contextmenu-01-open.png'), fullPage: true });
 
     expect(testLogger.errors.filter((e) => !e.includes('favicon'))).toHaveLength(0);
   });

tests/e2e/coroutine-nested.spec.ts

@@ -1,4 +1,5 @@
 import { test, expect, tryLoadApp } from './utils/fixtures';
+import { shotPath } from './utils/element-tracker';
 
 const EXPECTED_CASES = [
   'baseline_modal_alone',
@@ -36,7 +37,7 @@ test.describe('Nested Coroutine+Modal Tests', () => {
       ).toBe(true);
     }
 
-    await page.screenshot({ path: 'test-results/coroutine-nested-01-loaded.png', fullPage: true });
+    await page.screenshot({ path: shotPath(page, 'coroutine-nested-01-loaded.png'), fullPage: true });
 
     expect(loaded, 'Nested harness should load').toBe(true);
   });
@@ -81,7 +82,7 @@ test.describe('Nested Coroutine+Modal Tests', () => {
       'no unexpected page errors'
     ).toHaveLength(0);
 
-    await page.screenshot({ path: 'test-results/coroutine-nested-02-summary.png', fullPage: true });
+    await page.screenshot({ path: shotPath(page, 'coroutine-nested-02-summary.png'), fullPage: true });
   });
 
   test('per-scenario status (diagnostic)', async ({ page, testLogger }) => {

tests/e2e/coroutine.spec.ts

@@ -1,4 +1,5 @@
 import { test, expect, tryLoadApp } from './utils/fixtures';
+import { shotPath } from './utils/element-tracker';
 
 const EXPECTED_CASES = [
   'first_entry_runs_once',
@@ -36,7 +37,7 @@ test.describe('Coroutine Harness Tests', () => {
       expect(caseLogs.some((log) => log.includes(`[COROUTINE_TEST] CASE ${caseName}`))).toBe(true);
     }
 
-    await page.screenshot({ path: 'test-results/coroutine-01-loaded.png', fullPage: true });
+    await page.screenshot({ path: shotPath(page, 'coroutine-01-loaded.png'), fullPage: true });
 
     expect(loaded, 'Coroutine harness should load').toBe(true);
   });
@@ -69,6 +70,6 @@ test.describe('Coroutine Harness Tests', () => {
     expect(passLogs).toHaveLength(EXPECTED_CASES.length);
     expect(testLogger.errors.filter((error) => !error.includes('favicon'))).toHaveLength(0);
 
-    await page.screenshot({ path: 'test-results/coroutine-02-summary.png', fullPage: true });
+    await page.screenshot({ path: shotPath(page, 'coroutine-02-summary.png'), fullPage: true });
   });
 });

tests/e2e/retinascale.spec.ts

@@ -1,5 +1,6 @@
 // Retina/HiDPI Scaling Test - Verifies bitmaps render at correct logical pixel sizes
 import { test, expect, tryLoadApp } from './utils/fixtures';
+import { shotPath } from './utils/element-tracker';
 
 test.describe('Retina Scale Tests', () => {
 
@@ -7,7 +8,7 @@ test.describe('Retina Scale Tests', () => {
     await page.goto('/standalone/retinascale/retinascale_test.html');
     const loaded = await tryLoadApp(page);
 
-    await page.screenshot({ path: 'test-results/retinascale-01-loaded.png', fullPage: true });
+    await page.screenshot({ path: shotPath(page, 'retinascale-01-loaded.png'), fullPage: true });
 
     expect(loaded, 'Retina scale test app should load').toBe(true);
     expect(testLogger.errors.filter(e => !e.includes('favicon'))).toHaveLength(0);

tests/e2e/threadpool.spec.ts

@@ -1,4 +1,5 @@
 import { test, expect, tryLoadApp } from './utils/fixtures';
+import { shotPath } from './utils/element-tracker';
 
 test.describe('Thread Pool Deadlock Tests', () => {
 
@@ -17,7 +18,7 @@ test.describe('Thread Pool Deadlock Tests', () => {
     // If there's a deadlock, tryLoadApp will timeout waiting for canvas
     const loaded = await tryLoadApp(page, 30000);
 
-    await page.screenshot({ path: 'test-results/threadpool-01-loaded.png', fullPage: true });
+    await page.screenshot({ path: shotPath(page, 'threadpool-01-loaded.png'), fullPage: true });
 
     // Check for success marker - all threads created and joined
     const success = testLogger.consoleLogs.some(log =>

tests/e2e/utils/element-tracker.ts

@@ -1,4 +1,5 @@
-import { Page } from '@playwright/test';
+import { Page, Locator, expect as baseExpect } from '@playwright/test';
+import * as fs from 'fs';
 
 export interface WxElement {
   id: string;
@@ -1923,21 +1924,63 @@ export async function waitForRenderStable(page: Page, options: StableShotOptions
 }
 
 /**
- * Stabilize the render, then write a raw PNG to test-results/<name> for the offline screenshot gate
- * (tools/screenshots/compare.ts vs tests/baseline-screenshots/). Drop-in replacement for
- * `expect(page).toHaveScreenshot(name, opts)`: keeps toHaveScreenshot's render-settle but NOT its
- * inline comparison (that's the calibrated offline pipeline's job) nor its per-platform baseline
- * files. Non-asserting — a render regression is caught by `npm run screenshots:check`, not the test.
+ * Engine-scoped path for a gate screenshot: test-results/<engine>/<name>, where <engine> is the
+ * running browser family (chromium | firefox | webkit). Every engine gets its OWN namespace so a
+ * spec running on two browsers in one invocation can't have the engines overwrite each other's
+ * PNG (the old flat test-results/<name> made the surviving file whichever worker wrote last).
+ * The offline gate (tools/screenshots) compares each engine against its own committed baselines
+ * in tests/baseline-screenshots/<engine>/.
+ */
+export function shotPath(page: Page, name: string): string {
+  const engine = page.context().browser()!.browserType().name();
+  const dir = `test-results/${engine}`;
+  // page.screenshot creates parent dirs itself, but fs writers (canvas dumps) don't.
+  fs.mkdirSync(dir, { recursive: true });
+  return `${dir}/${name}`;
+}
+
+/**
+ * Stabilize the render, then write a raw PNG to test-results/<engine>/<name> for the offline
+ * screenshot gate (tools/screenshots/compare.ts vs tests/baseline-screenshots/<engine>/). Drop-in
+ * replacement for `expect(page).toHaveScreenshot(name, opts)`: keeps toHaveScreenshot's
+ * render-settle but NOT its inline comparison (that's the calibrated offline pipeline's job) nor
+ * its per-platform baseline files. Non-asserting — a render regression is caught by
+ * `npm run screenshots:check`, not the test. A name containing '/' is used verbatim (for writers
+ * outside the gate, e.g. gal-regression outputs).
  */
 export async function stableShot(page: Page, name: string, options: StableShotOptions = {}): Promise<void> {
   await waitForRenderStable(page, options);
   await page.screenshot({
-    path: name.includes('/') ? name : `test-results/${name}`,
+    path: name.includes('/') ? name : shotPath(page, name),
     fullPage: options.fullPage ?? false,
     ...SHOT_OPTS,
   });
 }
 
+/**
+ * Byte-stable screenshot of a locator (or the whole page): capture only once two consecutive
+ * screenshots are identical — the same equality a later "pixels restored" poll asserts, so
+ * before/after comparisons are self-consistent. Guards against baselining a mid-convergence
+ * frame the canvas never shows again. Pass the Page when the settled region must include
+ * wx dialogs/popups that render outside #canvas. (Shared by the presence overlay specs and
+ * the web gesture/chooser specs.)
+ */
+export async function settledShot(canvas: Locator | Page, expectImpl: typeof baseExpect = baseExpect): Promise<Buffer> {
+  let prev = await canvas.screenshot();
+  await expectImpl
+    .poll(
+      async () => {
+        const next = await canvas.screenshot();
+        const settled = next.equals(prev);
+        prev = next;
+        return settled;
+      },
+      { timeout: 30000, intervals: [250] },
+    )
+    .toBe(true);
+  return prev;
+}
+
 // ============================================================================
 // Label normalization + app-readiness (shared by the kicad and e2e/widget suites)
 // ============================================================================

tests/e2e/wxwidgets.spec.ts

@@ -10,7 +10,7 @@ import {
   clickTextCtrl,
   findSingleLineTextCtrl,
   findMultiLineTextCtrl,
-  clickListboxItemByIndex, stableShot } from './utils/element-tracker';
+  clickListboxItemByIndex, stableShot, shotPath } from './utils/element-tracker';
 
 // Capture console events with [EVENT] prefix
 function captureEvents(page: Page): string[] {
@@ -69,7 +69,7 @@ test.describe('wxWidgets WASM - Diagnostics', () => {
 
     // Loading-state capture: raw + immediate (no settle) on purpose — this shot documents the
     // app mid-load, before the canvas is up, so it must NOT wait for the render to finish.
-    await page.screenshot({ path: 'test-results/01-loading.png', fullPage: true });
+    await page.screenshot({ path: shotPath(page, '01-loading.png'), fullPage: true });
 
     // Deterministic app readiness: canvas visible + wx registry populated.
     // Fails loudly (replaces the try/catch waitForSelector + 1s settle).