style: rustfmt the two files #1116 left unformatted

[justinchuby]

Fast (Linux x86_64) is red on main for every open PR: #1116 landed with rustfmt drift.

Diff in crates/onnx-runtime-ep-cpu/src/kernels/matmul.rs:4950
Diff in crates/onnx-runtime-ep-cpu/src/kernels/x86_sgemm.rs:349

Verified on a pristine detached checkout of origin/main @ 1f1ce4b74, so it is not introduced by any open branch. This is pure cargo fmt --all output — no semantic change, no test change.

[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/small_generic_f32_threads=8/1x256x256	39.19 µs	67.80 µs	+73.0%
🔴	block_quantized_matmul_cached_dense/mxfp4_cached_dense_repeated_call/1x1024x1024	55.78 µs	86.71 µs	+55.5%
🔴	gather/large_bf16_threads=1-internal/131072	15.97 µs	21.04 µs	+31.8%
🔴	block_quantized_matmul_cached_dense/mxfp4_preexpanded_dense_oncelock_like_proxy/1x1024x1024	67.02 µs	88.26 µs	+31.7%
⚠️	matmul/small_generic_f32_threads=1/1x256x256	42.19 µs	53.63 µs	+27.1%
⚠️	matmul/medium_generic_f16_threads=1/32x512x512	32.47 µs	39.65 µs	+22.1%
⚠️	qwen3_sampling_processors/top_k_partial_selection	141.37 µs	168.64 µs	+19.3%
⚠️	qwen3_sampling_processors/top_k_full_sort_baseline	2.15 ms	2.56 ms	+19.1%
⚠️	tokenization/encode_tokens_per_second	377.16 µs	440.32 µs	+16.7%
✅	block_quantized_moe_cached_dense/mxfp4_uncached_expert_dequant_each_call/rows=1,H=256,I=256,E=4,top_k=1	594.71 µs	670.31 µs	+12.7%
✅	matmul/medium_generic_bf16_threads=1/32x512x512	561.35 µs	630.68 µs	+12.4%
✅	grammar_masking/llguidance_compute_mask/32	74.73 µs	83.25 µs	+11.4%
✅	logit_processing/seven_processor_chain_per_step	323.45 µs	357.50 µs	+10.5%
✅	qwen3_sampling_processors/top_p_full_sort_after_top_k_baseline	3.49 ms	3.83 ms	+9.6%
✅	matmul/medium_generic_f16_threads=8/32x512x512	33.20 µs	36.37 µs	+9.5%
✅	kv_cache/alloc_dealloc_pages	39.42 µs	43.05 µs	+9.2%
✅	matmul/small_generic_f16_threads=8/1x256x256	36.34 µs	39.05 µs	+7.5%
✅	qwen3_sampling_processors/top_k_top_p_fast	669.66 µs	717.96 µs	+7.2%
✅	qwen3_sampling_processors/top_p_fast_after_top_k	520.69 µs	556.95 µs	+7.0%
✅	matmul/large_generic_f16_threads=1/32x1024x1024	81.81 µs	86.43 µs	+5.7%
✅	tokenization/decode_tokens_per_second	6.10 ms	6.32 ms	+3.6%
✅	qwen3_sampling_processors/top_k_top_p_full_sort_baseline	6.00 ms	6.22 ms	+3.6%
✅	matmul/small_generic_bf16_threads=8/1x256x256	40.23 µs	40.82 µs	+1.5%
✅	block_quantized_moe_cached_dense/mxfp4_cached_dense_expert_repeated_call/rows=1,H=256,I=256,E=4,top_k=1	190.93 µs	191.61 µs	+0.4%
✅	matmul/small_generic_bf16_threads=1/1x256x256	37.91 µs	37.59 µs	-0.9%
✅	sampling_latency/greedy_per_token	3.25 µs	3.22 µs	-1.1%
✅	sampling_latency/top_k_per_token	55.31 µs	53.29 µs	-3.7%
✅	reduce_mean/large_f32_threads=1-internal/262144	1.16 ms	1.09 ms	-5.3%
✅	matmul/large_generic_bf16_threads=1/32x1024x1024	2.49 ms	2.33 ms	-6.4%
✅	gather/medium_bf16_threads=1-internal/32768	2.91 µs	2.67 µs	-8.2%
✅	matmul/medium_generic_bf16_threads=8/32x512x512	499.85 µs	457.81 µs	-8.4%
✅	matmul/large_generic_f32_threads=1/32x1024x1024	11.12 ms	10.16 ms	-8.6%
✅	matmul/medium_generic_f32_threads=1/32x512x512	2.69 ms	2.45 ms	-8.9%
✅	sampling_latency/min_p_per_token	237.50 µs	214.55 µs	-9.7%
✅	gather/small_f32_threads=1-internal/4096	829.1 ns	746.3 ns	-10.0%
✅	matmul/small_generic_f16_threads=1/1x256x256	39.19 µs	35.20 µs	-10.2%
✅	sampling_latency/top_p_per_token	449.62 µs	402.40 µs	-10.5%
✅	matmul/large_generic_f32_threads=8/32x1024x1024	5.69 ms	5.06 ms	-11.0%
✅	reduce_mean/medium_f32_threads=1-internal/65536	341.20 µs	303.15 µs	-11.2%
✅	gather/small_bf16_threads=1-internal/4096	601.6 ns	525.8 ns	-12.6%
✅	gather/small_f16_threads=1-internal/4096	626.5 ns	532.8 ns	-14.9%
🟢	gather/large_f16_threads=1-internal/131072	17.17 µs	14.30 µs	-16.7%
🟢	reduce_mean/small_f32_threads=1-internal/4096	20.20 µs	16.73 µs	-17.2%
🟢	matmul/large_generic_f16_threads=8/32x1024x1024	117.03 µs	94.00 µs	-19.7%
🟢	block_quantized_matmul_cached_dense/mxfp4_uncached_dequant_each_call/1x1024x1024	1.26 ms	1.00 ms	-20.6%
🟢	gather/large_f32_threads=1-internal/131072	37.64 µs	28.85 µs	-23.3%
🟢	add/large_bf16_threads=1-internal/4194304	2.38 ms	1.78 ms	-25.1%
🟢	add/large_f16_threads=1-internal/4194304	2.26 ms	1.64 ms	-27.4%
🟢	gather/medium_f16_threads=1-internal/32768	3.78 µs	2.70 µs	-28.5%
🟢	add/medium_bf16_threads=1-internal/262144	146.82 µs	103.64 µs	-29.4%
🟢	add/large_f32_threads=1-internal/4194304	945.90 µs	657.65 µs	-30.5%
🟢	add/medium_f16_threads=1-internal/262144	149.66 µs	103.51 µs	-30.8%
🟢	add/medium_f32_threads=1-internal/262144	38.81 µs	24.94 µs	-35.7%
🟢	matmul/large_generic_bf16_threads=8/32x1024x1024	2.61 ms	1.55 ms	-40.4%
🟢	matmul/medium_generic_f32_threads=8/32x512x512	1.74 ms	1.03 ms	-41.1%
🟢	add/small_bf16_threads=1-internal/1024	771.9 ns	451.8 ns	-41.5%
🟢	add/small_f32_threads=1-internal/1024	350.1 ns	193.7 ns	-44.7%
🟢	add/small_f16_threads=1-internal/1024	814.2 ns	448.1 ns	-45.0%
🟢	gather/medium_f32_threads=1-internal/32768	7.48 µs	3.91 µs	-47.7%

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: { 5.54 4.15 6.98 }

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)