feat(hipsr): add expand LLVM lowering

[mingyueliuh]

Summary

• lower hipsr.expand to the wrap_expand runtime call
• preserve AS1 input/output pointers while keeping requested shapes and generated shape arrays in AS0
• support both host shape operands and constant shape_attr forms
• register the lowering and add focused LIT coverage

Test plan

• Rebuild hip-mlir-opt
• Run the complete MorphizenMLIR LIT suite (356 passed, 2 unsupported)

AI assistance

Cursor assisted with isolating the existing Expand lowering changes, resolving registration conflicts against the latest main branch, and preparing tests and PR text. The changes were reviewed against the generated LLVM IR and validated by rebuilding hip-mlir-opt and running the complete LIT suite.

[github-actions]

Thanks for opening a PR!

This project follows LLVM's incremental-development and AI-tool-use
guidance. See CONTRIBUTING.md
for the project workflow.

Before requesting review, please check that:

1. The change is focused. Substantial work links the relevant issue
   or design discussion.
2. The PR documents relevant test results and updates affected
   documentation.
3. If AI tools provided substantial assistance, the description
   explains what was assisted and how it was validated, and commit
   trailers identify the tool. The contributor has reviewed and
   understands the result.

Reviewers are assigned through
CODEOWNERS where ownership
is configured.

[github-actions]

L2 Accuracy Results (EP vs CPU)

Model	Combined L2	Total Elems	Skipped NaN/Inf
conv_test_hybrid	4.8668E-07	64	0
GroupQueryAttention_seq256	25.2366	2621440	0
MatMulNBits_o_seq128	259.906	368640	0
QMoE_seq128	34.957	368640	0

Threshold: 0.01 | Run: 3449 - Commit: f97cff9

[github-actions]

MorphiZen EP Performance Results

Model	QPS	Session (s)	1st Infer (ms)	CPU%	Mem (MB)
full_model_seq128	7.50	6.38	359	3	1243
GroupQueryAttention_seq128	4373.51	1.70268	10	6	310
matmul_down_seq128	526.85	2.38	71	3	351

EPContext Export Performance

Model	QPS	Session (s)	1st Infer (ms)	CPU%	Mem (MB)
full_model_seq128	7.53	46.27	363	3	15590

EPContext Import Performance

Model	QPS	Session (s)	1st Infer (ms)	CPU%	Mem (MB)
full_model_seq128	7.52	9.80	366	3	15760

OGA Benchmark Results

Model	Warmup	Reps	Prompt Len	Gen Tokens	TTFT (ms)	TPS	Peak Mem (GB)	GPU Mem (GB)
gpt-oss-20b-webgpu-int4-rtn-block-32	1	5	128	128	179.7	79.5	1.33	13.54
Llama-3.1-8B-awq-g128-int4-asym-fp16-onnx-dml	1	5	128	128	253.7	40.9	1.22	6.43

OGA Wheel Smoke (Python benchmark_e2e.py)

Model	TTFT (ms)	TPS
Llama-3.1-8B-awq-g128-int4-asym-fp16-onnx-dml	196	39.7

Run: 3449 - Commit: f97cff9