feat(signals,solid,web): error hooks name where an error was thrown apart from where it was met

[ryansolid]

A C6 semantics refinement, surfaced while building the Sentry adapter against the linked build: both error hooks (and SSR_RENDER_ERROR_CONTAINED) reported only the boundary's owner path, so every component under one <Errored> grouped into one path — too coarse for issue grouping in exactly the tier the hook exists for (minified production, no source maps).

The two facts

• ownerPath — where it was thrown. Labels root-first up the owner chain of the computation that threw: component labels and named primitives (the compiler's inner memos ride along by their default name). Falls back to the boundary's chain when the throw crossed nothing the runtime could name.
• boundaryPath — where it was met (new). The same labels up the chain of the <Errored> that rendered its fallback (client and server), or the <Loading> that shipped the rejection (server, handling: "client"). What broke, and what the user saw.

<Loading> needed no special rule: on the client it doesn't catch (the rejection passes through to the nearest <Errored>, which is the met point); on the server it's the boundary whose handling the context already names.

Mechanics

• Client — no new bookkeeping. The engine wraps the first landing in a StatusError whose source is the thrower, and keeps that instance as the error propagates; CollectionQueue.notify already had it. reportClientError takes it as a third argument.
• Server — runWithOwner and the memo pull's inlined twin stamp the error into a WeakMap<object, Owner> as it escapes (first wins, object errors only, NotReadyError skipped). reportServerError reads it. Component bodies, effects and async attempts all run through those scopes, so ["<App>", "<Loading>", "<Child>"] comes out for the fragment case without a per-site change.
• Finding — SSR_RENDER_ERROR_CONTAINED is located at the throw; data.boundaryPath names the boundary beside data.boundary.

The "boundary" record is unchanged: it is about the boundary, and joins the finding by id.

Verification

signals 2760 · solid 623 + 825 types · web 1001 + 185 types. Hook specs on both sides and the server finding specs assert the split (<App> › <Errored> › <Bad> thrown / <App> › <Errored> met). One cap: hydrating 19.85 → 19.90 KB, +41 B for the second path over a baseline that sat 1 B under its cap; the other scenarios moved within noise. RFC 03/08/12 updated; changeset for @solidjs/signals, solid-js, @solidjs/web.

Consumer side: the @sentry/solid-2 draft (in sentry-javascript, unpublished until an rc carries this) tags solid.owner and solid.boundary from the pair.

[changeset-bot]

🦋 Changeset detected

Latest commit: 038470f

The changes in this PR will be included in the next version bump.

This PR includes changesets to release 11 packages

Name	Type
@solidjs/signals	Patch
solid-js	Patch
@solidjs/web	Patch
test-integration	Patch
@solidjs/element	Patch
@solidjs/h	Patch
@solidjs/html	Patch
@solidjs/universal	Patch
@solidjs/babel-plugin	Patch
@solidjs/compiler	Patch
@solidjs/diagnostics	Patch

Not sure what this means? Click here to learn what changesets are.

Click here if you're a maintainer who wants to add another changeset to this PR

[coveralls]

Coverage Report for CI Build 35260411368

Coverage remained the same at 71.304%

Details

• Coverage remained the same as the base build.
• Patch coverage: No coverable lines changed in this PR.
• No coverage regressions found.

Uncovered Changes

No uncovered changes found.

Coverage Regressions

No coverage regressions found.

---

Coverage Stats

Relevant Lines:	1035
Covered Lines:	783
Line Coverage:	75.65%
Relevant Branches:	798
Covered Branches:	524
Branch Coverage:	65.66%
Branches in Coverage %:	Yes
Coverage Strength:	15.06 hits per line

---

💛 - Coveralls

[codspeed]

Merging this PR will not alter performance

✅ 172 untouched benchmarks
⏩ 3 skipped benchmarks¹

---

_{Comparing error-thrower-path (038470f) with next (5a0a5f2)}

Footnotes

1. 3 benchmarks were skipped, so the baseline results were used instead. If they were deleted from the codebase, click here and archive them to remove them from the performance reports. ↩