style: apply rustfmt to matmul_nbits.rs

[justinchuby]

main is currently red on cargo fmt --check: #1101 landed one slice expression that stable rustfmt reformats. That fails the "Check formatting" step in both Fast (Linux x86_64) and Rust quality, on main and on every open PR.

This applies exactly the diff CI printed, verbatim:

             for c in 0..group {
-                let block_values = &packed_rows[c]
-                    [block * packed_block_size..(block + 1) * packed_block_size];
+                let block_values =
+                    &packed_rows[c][block * packed_block_size..(block + 1) * packed_block_size];

No behaviour change, no other file touched. cargo fmt --all -- --check is clean locally after it.

Co-authored-by: Copilot 223556219+Copilot@users.noreply.github.com

[codecov]

Codecov Report

❌ Patch coverage is 0% with 2 lines in your changes missing coverage. Please review.
✅ Project coverage is 79.92%. Comparing base (1158698) to head (df59a39).
⚠️ Report is 1 commits behind head on main.

Files with missing lines	Patch %	Lines
...es/onnx-runtime-ep-cpu/src/kernels/matmul_nbits.rs	0.00%	2 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #1108      +/-   ##
==========================================
+ Coverage   79.83%   79.92%   +0.09%     
==========================================
  Files         367      369       +2     
  Lines      157237   160217    +2980     
  Branches   157237   160217    +2980     
==========================================
+ Hits       125537   128061    +2524     
- Misses      26987    27425     +438     
- Partials     4713     4731      +18     

Flag	Coverage Δ
cli-ort-linux	83.79% <ø> (ø)
cli-ort-windows	83.40% <ø> (ø)
mlas	84.50% <ø> (?)
offline	79.70% <0.00%> (+<0.01%)	⬆️

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	75.66% <0.00%> (ø)

... and 9 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
🔴	sampling_latency/top_p_per_token	384.67 µs	506.42 µs	+31.7%
⚠️	tokenization/encode_tokens_per_second	397.29 µs	504.12 µs	+26.9%
⚠️	sampling_latency/top_k_per_token	52.94 µs	66.50 µs	+25.6%
⚠️	qwen3_sampling_processors/top_k_top_p_full_sort_baseline	5.99 ms	7.36 ms	+22.9%
⚠️	matmul/large_generic_bf16_threads=1/32x1024x1024	2.29 ms	2.80 ms	+22.2%
⚠️	tokenization/decode_tokens_per_second	6.60 ms	8.04 ms	+21.9%
⚠️	block_quantized_matmul_cached_dense/mxfp4_uncached_dequant_each_call/1x1024x1024	890.94 µs	1.08 ms	+21.2%
⚠️	qwen3_sampling_processors/top_k_full_sort_baseline	2.23 ms	2.66 ms	+19.3%
⚠️	sampling_latency/min_p_per_token	215.92 µs	256.87 µs	+19.0%
⚠️	qwen3_sampling_processors/top_k_partial_selection	156.75 µs	185.16 µs	+18.1%
⚠️	qwen3_sampling_processors/top_p_full_sort_after_top_k_baseline	3.94 ms	4.53 ms	+15.1%
✅	sampling_latency/greedy_per_token	3.31 µs	3.79 µs	+14.5%
✅	qwen3_sampling_processors/top_p_fast_after_top_k	535.35 µs	601.24 µs	+12.3%
✅	block_quantized_moe_cached_dense/mxfp4_cached_dense_expert_repeated_call/rows=1,H=256,I=256,E=4,top_k=1	187.44 µs	209.77 µs	+11.9%
✅	gather/large_bf16_threads=1-internal/131072	13.63 µs	14.82 µs	+8.7%
✅	logit_processing/seven_processor_chain_per_step	361.08 µs	392.19 µs	+8.6%
✅	grammar_masking/llguidance_compute_mask/32	87.12 µs	92.92 µs	+6.7%
✅	matmul/large_generic_f32_threads=8/32x1024x1024	5.45 ms	5.68 ms	+4.1%
✅	gather/large_f16_threads=1-internal/131072	14.30 µs	14.74 µs	+3.1%
✅	add/large_f16_threads=1-internal/4194304	1.71 ms	1.75 ms	+2.3%
✅	block_quantized_matmul_cached_dense/mxfp4_preexpanded_dense_oncelock_like_proxy/1x1024x1024	106.89 µs	108.97 µs	+1.9%
✅	matmul/medium_generic_f16_threads=8/32x512x512	47.15 µs	47.77 µs	+1.3%
✅	matmul/small_generic_f32_threads=1/1x256x256	42.37 µs	42.83 µs	+1.1%
✅	add/large_f32_threads=1-internal/4194304	777.88 µs	777.41 µs	-0.1%
✅	qwen3_sampling_processors/top_k_top_p_fast	741.80 µs	741.24 µs	-0.1%
✅	add/large_bf16_threads=1-internal/4194304	1.73 ms	1.72 ms	-0.3%
✅	matmul/medium_generic_f32_threads=8/32x512x512	1.45 ms	1.43 ms	-0.9%
✅	add/small_bf16_threads=1-internal/1024	511.9 ns	507.2 ns	-0.9%
✅	matmul/small_generic_f16_threads=1/1x256x256	40.02 µs	39.57 µs	-1.1%
✅	matmul/small_generic_bf16_threads=8/1x256x256	40.97 µs	40.26 µs	-1.7%
✅	reduce_mean/small_f32_threads=1-internal/4096	16.34 µs	15.98 µs	-2.2%
✅	matmul/large_generic_f32_threads=1/32x1024x1024	10.22 ms	9.89 ms	-3.2%
✅	block_quantized_moe_cached_dense/mxfp4_uncached_expert_dequant_each_call/rows=1,H=256,I=256,E=4,top_k=1	644.97 µs	618.60 µs	-4.1%
✅	gather/medium_f16_threads=1-internal/32768	2.78 µs	2.65 µs	-4.5%
✅	gather/small_f32_threads=1-internal/4096	727.1 ns	693.1 ns	-4.7%
✅	matmul/large_generic_f16_threads=1/32x1024x1024	90.43 µs	86.10 µs	-4.8%
✅	add/small_f16_threads=1-internal/1024	505.4 ns	480.1 ns	-5.0%
✅	matmul/small_generic_bf16_threads=1/1x256x256	40.60 µs	38.36 µs	-5.5%
✅	add/medium_bf16_threads=1-internal/262144	119.10 µs	112.48 µs	-5.6%
✅	reduce_mean/large_f32_threads=1-internal/262144	1.11 ms	1.04 ms	-6.0%
✅	gather/medium_f32_threads=1-internal/32768	4.47 µs	4.20 µs	-6.0%
✅	matmul/small_generic_f16_threads=8/1x256x256	41.40 µs	38.36 µs	-7.3%
✅	matmul/medium_generic_bf16_threads=1/32x512x512	592.27 µs	546.44 µs	-7.7%
✅	gather/small_f16_threads=1-internal/4096	536.7 ns	491.6 ns	-8.4%
✅	gather/small_bf16_threads=1-internal/4096	528.9 ns	484.0 ns	-8.5%
✅	matmul/medium_generic_f16_threads=1/32x512x512	40.70 µs	36.67 µs	-9.9%
✅	add/medium_f16_threads=1-internal/262144	119.75 µs	107.80 µs	-10.0%
✅	add/medium_f32_threads=1-internal/262144	28.26 µs	25.41 µs	-10.1%
✅	gather/large_f32_threads=1-internal/131072	41.76 µs	37.32 µs	-10.6%
✅	matmul/large_generic_f16_threads=8/32x1024x1024	113.77 µs	101.12 µs	-11.1%
✅	matmul/medium_generic_f32_threads=1/32x512x512	2.73 ms	2.42 ms	-11.2%
✅	matmul/small_generic_f32_threads=8/1x256x256	58.74 µs	51.98 µs	-11.5%
✅	reduce_mean/medium_f32_threads=1-internal/65536	284.29 µs	249.97 µs	-12.1%
✅	kv_cache/alloc_dealloc_pages	47.28 µs	40.79 µs	-13.7%
🟢	block_quantized_matmul_cached_dense/mxfp4_cached_dense_repeated_call/1x1024x1024	123.77 µs	98.81 µs	-20.2%
🟢	matmul/medium_generic_bf16_threads=8/32x512x512	656.50 µs	519.38 µs	-20.9%
🟢	matmul/large_generic_bf16_threads=8/32x1024x1024	2.43 ms	1.87 ms	-23.0%
🟢	gather/medium_bf16_threads=1-internal/32768	3.37 µs	2.41 µs	-28.5%
🟢	add/small_f32_threads=1-internal/1024	283.9 ns	200.5 ns	-29.4%

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: { 4.70 4.48 5.45 }

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)