qwen3_5: retain multimodal Graph regressions in workspace tests

[Levius-Fubuki]

Purpose

Retain native multimodal language CUDA Graph regressions under root tests/, addressing #103’s maintenance review. Three opt-in GPU cases and the capture-failure runner cover ordered cache keys, current IDs/images/positions, multimodal interleave, scratch growth, FIFO eviction and eager fallback.

Depends on #103. The main-based diff includes its production implementation; this PR’s independent addition is the regression suite and four lines of private test-module wiring.

Submission scope

Core implementation and directly related tests are retained. Ordinary documentation, experiment reports/media and generated run outputs remain outside this submission. Test fixtures, build inputs, deployment data and third-party licensing remain as needed.

Validation

After cleaning workspace debug artifacts before each branch, fresh local fmt, Clippy/all-targets with warnings denied and 90 Rust tests pass (16 opt-in tests ignored). Independent GitHub CI passes on this exact head, including release workspace build and strict Docs. Restored test bodies/fixtures match the saved originals. Production execution is unchanged. GPU tests were not rerun; the server remains shut down.

Head 968a9be2763d0cf79a8935b89c493f562ded92e2; 3 changed files and 532 changed source/test/build lines, excluding licenses, notices, locks and static data.

Self-review

Re-reviewed complete test-only commit and parent integration against current contributor/architecture rules. No generated run output is committed. Full agent-assisted self-review is complete under the author’s explicit Ready instruction. Retain the #103 dependency and rebase after that parent merges.

• I have reviewed the full diff and addressed the issues I found.
• I have checked that the change follows the project's architecture and stays focused on the stated purpose.
• I have run the checks appropriate to this change and reported commands, results, and anything I could not verify above.
• I have checked that the PR description, documentation, and any accuracy or performance claims match the implementation and available evidence.

2026-10-08 verification follow-up

Validated source head: 2dd88db234b99a64061dbfd0094a4ab6b0b4e5cb. Upstream main snapshot: 4a79980d; GitHub reports no merge conflict. Fresh exact-head fmt, strict workspace Clippy, workspace tests and release build pass: 90 passed, 0 failed, 16 ignored. Ignored tests are counted separately from explicit GPU execution. Exact commands and logs are in the archive. Exact test head 2dd88db2: 3/3 release GPU cases plus both unconditional failure diagnostics pass on RTX 4090 with tracing unset. No assertion or tolerance changed. This closes the current-head GPU execution gap. Merge #103 first, then rebase to leave the dedicated test increment.

Raw verification evidence, SHA-256 5df66bf3974d534875946168fba4d7fe543d08edbd658c3756f17d6f0c8f5773. Includes exact source manifests, environment/library/binary hashes, complete logs, synthetic reference/HTTP results and initial setup failures. CUDA execution is serial on one RTX 4090. No weights, credentials or private inputs are included. Contributor/maintainer review and dependency merges remain separate.

Full self-review completed / Ready — 2026-10-08

Current head: 2dd88db234b99a64061dbfd0094a4ab6b0b4e5cb; merge base: 99865743d27316fbe81362dc0f0e6de6fda86284. Full cumulative diff: 532 authored-code lines / 579 total diff lines. The author explicitly requested completed self-review and Ready conversion. The completed checklist records this agent-assisted review and its verification; it does not represent maintainer approval.

Root integration review plus three independent component reviews covered all changed source, tests, helpers, manifests and documentation: CPU contract/checkpoint/tokenizer/calibration; serial worker/admission/GPU head and error retirement; ordered bounded packing and response reconstruction; Graph lifetime/cache/health/options; ABI7 CUDA continuations/fixed GEMM/request-local prefix state; root regressions, recipe reproduction, fixture consumers and evidence/claim boundaries. No functional blocker remains. For #112/#113 above 3,000 authored lines, this is full component/integration self-review, with the existing split rationale and review order retained.

Corrected stale unsupported-Graph and no-worker-prefix-consumer documentation where applicable. Removed unused CPU fixture run/duplicate fields on the Decider stack, preserving all 255 label IDs and all 12 used fixture values exactly; the original removed data remain in evidence outside committed source. Both pinned CPU contract opt-ins pass on every repaired Decider branch; repaired docs pass strict MkDocs. All five initial review heads passed fresh fmt, strict workspace Clippy, workspace tests, release build and strict docs; latest-head CI Rust/benchmark/docs also passes. Ignored hardware/model tests are separate from executed tests.

Runtime/CUDA/GPU regression code is unchanged by this cleanup. Earlier personally generated GPU/reference/HTTP campaigns retain their original source revisions; independent reviewers rehashed both source bindings, replayed all stored numerical gates and fixed/shared exact controls, and recomputed published timings. ready-source-binding.json binds these runtime checks to the latest heads and separately proves the used CPU oracle values are unchanged. No new CUDA/HTTP/performance campaign was run after server shutdown; the server remains off. Existing hardware, scope and performance limitations remain disclosed.

Complete self-review reports and fresh checks; SHA-256 1edbd4a6e13576d4cc23638b7ec1e35c1248d6a712b8a6406e1f269ca685cb9d. Includes all four review reports, closure summary, original reviewed full diffs, new CPU/docs logs, latest-head CI, source/fixture binding and archived removed fields. Merge dependencies remain #103 before #106, and #94 → #110 → #111 → #112 → #113, including #97/#98/#99 prerequisites for #113; Ready conversion does not merge them.

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.