style: rustfmt matmul_nbits, which is failing CI on main

[justinchuby]

main fails Check formatting in both the Fast (Linux x86_64) and Rust quality jobs (run 32069727921), so every open PR inherits a red CI through the merge commit.

This is cargo fmt --all output and nothing else — an over-indented .expect(...) continuation and an eprintln! that fits on one line, both inside mod tests in crates/onnx-runtime-ep-cpu/src/kernels/matmul_nbits.rs.

$ git diff --stat
 crates/onnx-runtime-ep-cpu/src/kernels/matmul_nbits.rs | 24 ++++++++++--------
 1 file changed, 11 insertions(+), 13 deletions(-)

$ git diff -w --stat      # ignoring whitespace
 crates/onnx-runtime-ep-cpu/src/kernels/matmul_nbits.rs | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)   # one joined eprintln! line

No code path changes. Verified: cargo fmt --all -- --check clean, cargo check -p onnx-runtime-ep-cpu --tests builds.

The file belongs to Roy's #1138 / ARM MatMulNBits work; this touches only rustfmt output there and not that investigation.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 79.95%. Comparing base (c62798b) to head (f5d07d4).
⚠️ Report is 1 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #1144      +/-   ##
==========================================
- Coverage   80.47%   79.95%   -0.53%     
==========================================
  Files         368      368              
  Lines      160934   160931       -3     
  Branches   160934   160931       -3     
==========================================
- Hits       129517   128666     -851     
- Misses      26691    27543     +852     
+ Partials     4726     4722       -4     

Flag	Coverage Δ
cli-ort-linux	83.79% <ø> (ø)
cli-ort-windows	?
mlas	85.67% <ø> (-0.14%)	⬇️
offline	79.70% <ø> (-0.56%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

Files with missing lines	Coverage Δ
...es/onnx-runtime-ep-cpu/src/kernels/matmul_nbits.rs	75.91% <ø> (-5.73%)	⬇️

... and 8 files with indirect coverage changes

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[github-actions]

🔴 Benchmark Regression Detected

Comparison of criterion micro-benchmarks: PR head vs merge-base, measured on the same runner in the same job (base first → PR second).

ℹ️ Absolute times are informational only — they vary with runner load. The % change column is the reliable signal because both sides ran under identical conditions.

Status	Scenario	Base	PR	Change
🔴	matmul/medium_generic_f32_threads=8/32x512x512	953.23 µs	1.97 ms	+106.9%
🔴	block_quantized_matmul_cached_dense/mxfp4_uncached_dequant_each_call/1x1024x1024	499.62 µs	752.76 µs	+50.7%
🔴	matmul/small_generic_bf16_threads=8/1x256x256	30.26 µs	41.95 µs	+38.6%
🔴	matmul/medium_generic_f16_threads=8/32x512x512	27.62 µs	37.02 µs	+34.1%
🔴	kv_cache/alloc_dealloc_pages	36.23 µs	48.09 µs	+32.7%
🔴	matmul/medium_generic_f32_threads=1/32x512x512	2.15 ms	2.82 ms	+31.5%
🔴	block_quantized_matmul_cached_dense/mxfp4_cached_dense_repeated_call/1x1024x1024	40.54 µs	53.30 µs	+31.5%
🔴	matmul/large_generic_bf16_threads=8/32x1024x1024	1.28 ms	1.67 ms	+30.3%
⚠️	matmul/medium_generic_bf16_threads=8/32x512x512	371.58 µs	480.05 µs	+29.2%
⚠️	add/large_bf16_threads=1-internal/4194304	1.63 ms	2.05 ms	+25.8%
⚠️	logit_processing/seven_processor_chain_per_step	296.66 µs	370.50 µs	+24.9%
⚠️	matmul/medium_generic_f16_threads=1/32x512x512	28.65 µs	34.56 µs	+20.6%
⚠️	gather/large_f16_threads=1-internal/131072	10.65 µs	12.84 µs	+20.6%
⚠️	matmul/large_generic_f16_threads=8/32x1024x1024	78.09 µs	93.32 µs	+19.5%
⚠️	qwen3_sampling_processors/top_k_top_p_fast	607.08 µs	723.06 µs	+19.1%
⚠️	matmul/small_generic_f16_threads=8/1x256x256	28.29 µs	33.36 µs	+17.9%
⚠️	reduce_mean/small_f32_threads=1-internal/4096	14.75 µs	17.08 µs	+15.7%
✅	grammar_masking/llguidance_compute_mask/32	68.49 µs	78.69 µs	+14.9%
✅	gather/medium_bf16_threads=1-internal/32768	2.31 µs	2.55 µs	+10.5%
✅	qwen3_sampling_processors/top_k_top_p_full_sort_baseline	5.20 ms	5.60 ms	+7.6%
✅	qwen3_sampling_processors/top_k_partial_selection	129.17 µs	138.25 µs	+7.0%
✅	matmul/large_generic_f16_threads=1/32x1024x1024	81.05 µs	86.21 µs	+6.4%
✅	matmul/large_generic_f32_threads=1/32x1024x1024	8.65 ms	9.19 ms	+6.2%
✅	matmul/large_generic_bf16_threads=1/32x1024x1024	1.83 ms	1.95 ms	+6.1%
✅	gather/large_f32_threads=1-internal/131072	28.30 µs	29.93 µs	+5.8%
✅	add/large_f16_threads=1-internal/4194304	1.66 ms	1.74 ms	+5.2%
✅	reduce_mean/medium_f32_threads=1-internal/65536	249.69 µs	260.12 µs	+4.2%
✅	matmul/small_generic_bf16_threads=1/1x256x256	28.69 µs	29.85 µs	+4.0%
✅	add/large_f32_threads=1-internal/4194304	772.28 µs	797.78 µs	+3.3%
✅	block_quantized_moe_cached_dense/mxfp4_uncached_expert_dequant_each_call/rows=1,H=256,I=256,E=4,top_k=1	426.86 µs	437.62 µs	+2.5%
✅	matmul/medium_generic_bf16_threads=1/32x512x512	560.85 µs	572.54 µs	+2.1%
✅	qwen3_sampling_processors/top_k_full_sort_baseline	1.99 ms	2.02 ms	+1.5%
✅	qwen3_sampling_processors/top_p_full_sort_after_top_k_baseline	3.22 ms	3.27 ms	+1.5%
✅	sampling_latency/top_k_per_token	48.55 µs	49.21 µs	+1.4%
✅	tokenization/encode_tokens_per_second	348.77 µs	352.48 µs	+1.1%
✅	tokenization/decode_tokens_per_second	5.80 ms	5.84 ms	+0.7%
✅	gather/medium_f16_threads=1-internal/32768	2.55 µs	2.56 µs	+0.6%
✅	sampling_latency/min_p_per_token	193.30 µs	193.50 µs	+0.1%
✅	block_quantized_moe_cached_dense/mxfp4_cached_dense_expert_repeated_call/rows=1,H=256,I=256,E=4,top_k=1	154.61 µs	154.68 µs	+0.0%
✅	sampling_latency/top_p_per_token	357.56 µs	356.87 µs	-0.2%
✅	add/medium_bf16_threads=1-internal/262144	104.47 µs	104.22 µs	-0.2%
✅	matmul/small_generic_f16_threads=1/1x256x256	29.13 µs	28.80 µs	-1.1%
✅	matmul/large_generic_f32_threads=8/32x1024x1024	4.31 ms	4.25 ms	-1.4%
✅	gather/small_bf16_threads=1-internal/4096	517.3 ns	506.1 ns	-2.2%
✅	sampling_latency/greedy_per_token	3.05 µs	2.99 µs	-2.2%
✅	matmul/small_generic_f32_threads=8/1x256x256	32.29 µs	31.30 µs	-3.1%
✅	add/medium_f16_threads=1-internal/262144	106.53 µs	102.36 µs	-3.9%
✅	gather/small_f32_threads=1-internal/4096	703.1 ns	669.9 ns	-4.7%
✅	qwen3_sampling_processors/top_p_fast_after_top_k	526.18 µs	500.33 µs	-4.9%
✅	gather/medium_f32_threads=1-internal/32768	4.27 µs	4.02 µs	-5.9%
✅	add/medium_f32_threads=1-internal/262144	24.41 µs	22.62 µs	-7.3%
✅	add/small_f16_threads=1-internal/1024	457.1 ns	411.5 ns	-10.0%
✅	add/small_f32_threads=1-internal/1024	205.9 ns	181.3 ns	-11.9%
✅	add/small_bf16_threads=1-internal/1024	461.7 ns	403.8 ns	-12.5%
✅	reduce_mean/large_f32_threads=1-internal/262144	1.15 ms	1.00 ms	-13.0%
✅	gather/large_bf16_threads=1-internal/131072	13.01 µs	11.21 µs	-13.8%
✅	block_quantized_matmul_cached_dense/mxfp4_preexpanded_dense_oncelock_like_proxy/1x1024x1024	48.68 µs	41.39 µs	-15.0%
🟢	matmul/small_generic_f32_threads=1/1x256x256	41.58 µs	33.88 µs	-18.5%
🟢	gather/small_f16_threads=1-internal/4096	585.6 ns	476.0 ns	-18.7%

Visual flags: ⚠️ ≥ 15% slower, 🔴 ≥ 30% slower — calibrated against measured runner noise (~27% worst-case on multi-threaded matmul)

Host info

CPU: Apple M1 (Virtual)
Cores: 3
OS: Darwin 25.5.0 arm64
Rust: rustc 1.97.1 (8bab26f4f 2026-07-14)
Load avg: { 2.55 3.37 6.47 }

What this cannot catch

• Regressions in code paths not covered by these benchmarks (e.g., end-to-end decode with a real model)
• Sub-threshold regressions that compound over multiple PRs
• Performance changes that only manifest under GPU execution
• Latency changes in the ORT integration path (these benchmarks exercise the native Rust kernels)