fix(engine): restore rustfmt on main (#1200 left a trailing blank line)

[justinchuby]

cargo fmt --all --check fails on main as of #1200 (2f8ba9b88), which appended a trailing blank
line to the end of crates/onnx-genai-engine/src/pipeline/mod.rs:

Diff in crates/onnx-genai-engine/src/pipeline/mod.rs:3681:
         }
     }
 }
-

Rust quality is one of the two required checks on main, so until this lands every open PR in the
repository is blocked on a stray newline — including three of mine sitting on auto-merge.

The fix is the one-line deletion cargo fmt --all produces. Nothing else in the tree moved.

After: cargo fmt --all --check clean.

[justinchuby]

Checked this against current main and it no longer reproduces. crates/onnx-genai-engine/src/pipeline/mod.rs is 3702 lines and its last line is } -- there is no trailing blank line at 3681 -- and cargo fmt --all -- --check does not flag the file. The diff here is against an older main.

--check was red on main, but for a different reason: #1214 inserted an import without reflowing two use lists in the CLI. Fixed in #1217. Closing this as stale; reopen if you can still reproduce the blank line.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 79.90%. Comparing base (af9ee4e) to head (6bc3c62).
⚠️ Report is 20 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff            @@
##           main    #1204       +/-   ##
=========================================
+ Coverage      0   79.90%   +79.90%     
=========================================
  Files         0      359      +359     
  Lines         0   157892   +157892     
  Branches      0   157892   +157892     
=========================================
+ Hits          0   126162   +126162     
- Misses        0    27117    +27117     
- Partials      0     4613     +4613     

Flag	Coverage Δ
mlas	85.39% <ø> (?)
offline	79.79% <ø> (?)

Flags with carried forward coverage won't be shown. Click here to find out more.
see 359 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 Change 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
⚠️	gather/large_f16_threads=1-internal/131072	16.52 µs	19.47 µs	+17.9%
⚠️	gather/medium_bf16_threads=1-internal/32768	2.45 µs	2.89 µs	+17.6%
✅	gather/large_bf16_threads=1-internal/131072	14.87 µs	17.03 µs	+14.5%
✅	gather/small_f32_threads=1-internal/4096	671.9 ns	730.9 ns	+8.8%
✅	add/medium_f16_threads=1-internal/262144	106.63 µs	115.16 µs	+8.0%
✅	reduce_mean/large_f32_threads=1-internal/262144	1.00 ms	1.07 ms	+7.1%
✅	reduce_mean/small_f32_threads=1-internal/4096	16.49 µs	17.46 µs	+5.9%
✅	matmul/medium_generic_f32_threads=1/32x512x512	2.50 ms	2.63 ms	+5.0%
✅	matmul/medium_generic_bf16_threads=1/32x512x512	559.34 µs	577.75 µs	+3.3%
✅	add/large_f16_threads=1-internal/4194304	1.77 ms	1.83 ms	+3.1%
✅	gather/medium_f16_threads=1-internal/32768	2.55 µs	2.61 µs	+2.5%
✅	matmul/large_generic_bf16_threads=1/32x1024x1024	2.15 ms	2.20 ms	+2.5%
✅	add/medium_f32_threads=1-internal/262144	24.96 µs	25.43 µs	+1.9%
✅	add/large_bf16_threads=1-internal/4194304	1.70 ms	1.72 ms	+1.7%
✅	gather/large_f32_threads=1-internal/131072	38.54 µs	38.93 µs	+1.0%
✅	block_quantized_moe_cached_dense/mxfp4_cached_dense_expert_repeated_call/rows=1,H=256,I=256,E=4,top_k=1	175.18 µs	176.90 µs	+1.0%
✅	add/small_bf16_threads=1-internal/1024	453.1 ns	454.9 ns	+0.4%
✅	matmul/large_generic_f32_threads=1/32x1024x1024	10.02 ms	9.95 ms	-0.7%
✅	gather/medium_f32_threads=1-internal/32768	4.15 µs	4.12 µs	-0.8%
✅	reduce_mean/medium_f32_threads=1-internal/65536	255.93 µs	253.37 µs	-1.0%
✅	gather/small_f16_threads=1-internal/4096	514.1 ns	499.7 ns	-2.8%
✅	add/medium_bf16_threads=1-internal/262144	107.61 µs	103.21 µs	-4.1%
✅	gather/small_bf16_threads=1-internal/4096	492.6 ns	471.9 ns	-4.2%
✅	tokenization/decode_tokens_per_second	7.85 ms	7.45 ms	-5.1%
✅	matmul/medium_generic_f16_threads=1/32x512x512	38.37 µs	35.95 µs	-6.3%
✅	tokenization/encode_tokens_per_second	435.28 µs	407.12 µs	-6.5%
✅	sampling_latency/greedy_per_token	3.76 µs	3.47 µs	-7.7%
✅	qwen3_sampling_processors/top_k_top_p_full_sort_baseline	6.18 ms	5.67 ms	-8.3%
✅	add/small_f32_threads=1-internal/1024	236.8 ns	215.5 ns	-9.0%
✅	sampling_latency/top_p_per_token	441.79 µs	395.56 µs	-10.5%
✅	add/small_f16_threads=1-internal/1024	558.9 ns	500.1 ns	-10.5%
✅	sampling_latency/top_k_per_token	60.71 µs	54.31 µs	-10.5%
✅	matmul/large_generic_f16_threads=1/32x1024x1024	88.14 µs	78.81 µs	-10.6%
✅	logit_processing/seven_processor_chain_per_step	377.51 µs	336.74 µs	-10.8%
✅	qwen3_sampling_processors/top_k_top_p_fast	698.71 µs	612.32 µs	-12.4%
✅	matmul/small_generic_bf16_threads=1/1x256x256	38.12 µs	33.34 µs	-12.6%
✅	add/large_f32_threads=1-internal/4194304	763.78 µs	660.09 µs	-13.6%
✅	matmul/small_generic_f16_threads=8/1x256x256	38.61 µs	33.33 µs	-13.7%
✅	qwen3_sampling_processors/top_p_fast_after_top_k	562.20 µs	483.46 µs	-14.0%
✅	matmul/small_generic_f32_threads=1/1x256x256	46.34 µs	39.75 µs	-14.2%
✅	matmul/large_generic_f16_threads=8/32x1024x1024	108.60 µs	92.79 µs	-14.6%
✅	matmul/small_generic_f16_threads=1/1x256x256	38.60 µs	32.94 µs	-14.7%
✅	kv_cache/alloc_dealloc_pages	43.82 µs	37.25 µs	-15.0%
🟢	matmul/medium_generic_f16_threads=8/32x512x512	45.67 µs	37.87 µs	-17.1%
🟢	qwen3_sampling_processors/top_p_full_sort_after_top_k_baseline	4.02 ms	3.30 ms	-18.0%
🟢	qwen3_sampling_processors/top_k_partial_selection	160.47 µs	131.27 µs	-18.2%
🟢	matmul/small_generic_f32_threads=8/1x256x256	50.62 µs	41.07 µs	-18.9%
🟢	matmul/small_generic_bf16_threads=8/1x256x256	40.03 µs	32.47 µs	-18.9%
🟢	matmul/medium_generic_f32_threads=8/32x512x512	1.84 ms	1.39 ms	-24.6%
🟢	block_quantized_moe_cached_dense/mxfp4_uncached_expert_dequant_each_call/rows=1,H=256,I=256,E=4,top_k=1	638.93 µs	461.54 µs	-27.8%
🟢	matmul/large_generic_f32_threads=8/32x1024x1024	6.97 ms	4.82 ms	-30.8%
🟢	qwen3_sampling_processors/top_k_full_sort_baseline	3.05 ms	1.99 ms	-34.6%
🟢	matmul/medium_generic_bf16_threads=8/32x512x512	704.91 µs	437.06 µs	-38.0%
🟢	matmul/large_generic_bf16_threads=8/32x1024x1024	2.33 ms	1.33 ms	-42.9%
🟢	grammar_masking/llguidance_compute_mask/32	130.36 µs	70.87 µs	-45.6%
🟢	block_quantized_matmul_cached_dense/mxfp4_uncached_dequant_each_call/1x1024x1024	1.21 ms	655.43 µs	-45.7%
🟢	block_quantized_matmul_cached_dense/mxfp4_preexpanded_dense_oncelock_like_proxy/1x1024x1024	98.44 µs	51.55 µs	-47.6%
🟢	block_quantized_matmul_cached_dense/mxfp4_cached_dense_repeated_call/1x1024x1024	128.31 µs	66.38 µs	-48.3%
🟢	sampling_latency/min_p_per_token	455.85 µs	213.81 µs	-53.1%

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: { 3.63 4.45 5.56 }

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)