[BUG] The pass "fuse_pad_into_pool" causes output mismatch

The pass "fuse_pad_into_pool" causes output mismatch

Issue
Running the single pass fuse_pad_into_pool with onnxoptimizer 0.4.2 changes numerical outputs. One real model regresses after this pass only, while the original models run correctly with the same oracle inputs.

Environment

• Ubuntu 20.04
• Python 3.10
• onnx==1.19.0
• onnxruntime==1.19.2
• onnxoptimizer==0.4.2 (latest)

Repro steps (run from this folder)

1. Download and unzip the attached archive, then cd into the extracted directory

fuse_pad_into_pool_repro.tar.gz

tar -xzvf fuse_pad_into_pool_repro.tar.gz
cd fuse_pad_into_pool_repro

2. Create a Python 3.10 virtual environment and install dependencies:

   python -m venv .venv
   source .venv/bin/activate
   pip install --upgrade pip
   pip install -r requirements.txt
   

3. Optimize the case with only fuse_pad_into_pool (writes model.opt.onnx next to model.onnx):

   python optimize_model.py --case ./case_02937_seed43509122
   

4. Differential test original vs optimized outputs using stored oracle inputs (ORT graph optimizations disabled):

   python diff_test.py --case ./case_02937_seed43509122
   

Observed results

• case_02937_seed43509122: overall max_abs=1.221e-04, max_rel=9.506e-04; output drift observed.

Expected
fuse_pad_into_pool should be semantics-preserving. Applying this pass alone should not change any output values.

Differential Test Output Details

1. case_02937_seed43509122

Case: case_02937_seed43509122
Mismatched outputs (optimized vs reference):
  v17_0 (output[12]): max_abs=1.221e-04, max_rel=7.168e-04, shape=(2, 1, 16, 2)
  v49_0 (output[17]): max_abs=1.221e-04, max_rel=9.756e-04, shape=(44, 1, 16, 46, 2)
Overall: max_abs=1.221e-04, max_rel=9.756e-04

Attachments

• README.md
• requirements.txt
• optimize_model.py (runs only fuse_pad_into_pool and saves optimized as model.opt.onnx)
• diff_test.py (runs original and optimized model with same oracle inputs and reports max_abs/max_rel)
• run_optimized_model.py (runs the optimized model to inspect outputs)
• case_02937_seed43509122/ (original model with oracle inputs for reproduction)

Start with README.md and run optimize_model.py and diff_test.py in the attached fuse_pad_into_pool_repro archive to reproduce the reported drift. Inspect the single-pass optimization flow and the case_02937_seed43509122 model using the provided scripts. Done means the pass preserves the original outputs within the issue's differential test.