fix(ir): dual-dispatch manual no-split pipe groups

[lyfne123]

Summary

• infer dual_aiv_dispatch for hand-written Ascend910B AIC/AIV Groups that use no-split cross-core pipes
• preserve standalone behavior by creating Group-private adapters when a member is reused elsewhere
• assert orchestration emits {AIC, AIV, AIV} so lane 1 replays the required zero-valid-shape handshake

Follow-up to #2104 and #2097. The shared Group ABI fix removed the argument mismatch, but the original manual-pipe case still hung because the AIV member was dispatched only on lane 0.

Validation

• cmake --build build --parallel 8
• test_expand_mixed_kernel_a2a3.py: 20 passed
• issue regression in test_orchestration_tensor_rw.py: passed
• original unmodified probe_manual_pipe_draft.py -p a2a3 -d 1 N=8 device run: PASS in 3.22s (runtime 2.41s)

[gemini-code-assist]

Caution

The consumer version of Gemini Code Assist on GitHub has been sunset. All code review activity has officially ceased.

[coderabbitai]

Important

Review skipped

Auto incremental reviews are disabled on this repository.

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro Plus

Run ID: c4417740-85c6-4e07-a116-c416b0746e8c

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

📝 Walkthrough

Walkthrough

ExpandMixedKernel now infers dual_aiv_dispatch for eligible Ascend910B no-split cross-core pipe groups, independently of ABI normalization. Tests cover the inferred AIV attribute and generated mixed-kernel task ordering.

Changes

Dual AIV dispatch inference

Layer / File(s)	Summary
Infer and apply dispatch attributes src/ir/transforms/expand_mixed_kernel_pass.cpp	The pass detects supported no-split cross-core pipe usage, separates ABI normalization from dispatch inference, and conditionally applies dual_aiv_dispatch to the normalized AIV function.
Validate transformed orchestration tests/ut/ir/transforms/test_expand_mixed_kernel_a2a3.py, tests/ut/codegen/test_orchestration_tensor_rw.py	Regression coverage verifies inferred dispatch on a hand-written Group and checks the emitted mixed-kernel task IDs and AIV attribute.

Estimated code review effort: 3 (Moderate) | ~20 minutes

Sequence Diagram(s)

sequenceDiagram
  participant HandWrittenGroup
  participant ExpandMixedKernel
  participant AIVFunction
  participant OrchestrationCode
  HandWrittenGroup->>ExpandMixedKernel: provide AIC/AIV members with explicit A2A3 pipe
  ExpandMixedKernel->>AIVFunction: set dual_aiv_dispatch=True
  ExpandMixedKernel->>OrchestrationCode: emit MixedKernels mixed_0 with resolved task IDs

Loading

Possibly related PRs

• hw-native-sys/pypto#2009: Updates propagation of dual_aiv_dispatch=True for external mixed kernels and specialization behavior.

Poem

A rabbit hops through pipes of light,
AIV dispatch now takes flight.
Cube and vector, side by side,
March in their mixed-kernel stride.
Tests count each task just right.

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title concisely captures the main change: fixing dual-dispatch for manual no-split pipe groups.
Description check	✅ Passed	The description matches the changeset and describes the dual AIV dispatch fix and its validation.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: beabd5748f

ℹ️ About Codex in GitHub

Codex has been enabled to automatically review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

When you sign up for Codex through ChatGPT, Codex can also answer questions or update the PR, like "@codex address that feedback".