pytorch - 💡(How to fix) Fix torch.compile skips out-of-range check for torch.select and returns invalid values

🐛 Describe the bug

torch.compile with the Inductor backend does not preserve eager error behavior for torch.select when the selected index is out of range.

Eager execution correctly raises an IndexError, while the compiled version returns a tensor. The compiled output also contains invalid-looking values in the out-of-range part.

import torch

def f(x):
    base = x.unsqueeze(1)
    a = torch.select(base, dim=1, index=0)
    b = torch.select(base, dim=1, index=1)
    return torch.cat([a, b], dim=-1)

x = torch.arange(12, dtype=torch.float32).reshape(3, 4)

try:
    eager_out = f(x)
    eager = f"OK {eager_out}"
except Exception as exc:
    eager = f"{type(exc).__name__} {str(exc).splitlines()[0]}"

try:
    compiled_out = torch.compile(f, backend="inductor", fullgraph=True)(x)
    compiled = f"OK {compiled_out}"
except Exception as exc:
    compiled = f"{type(exc).__name__} {str(exc).splitlines()[0]}"

print("x:", x)
print("eager:", eager)
print("compiled:", compiled)

if not eager.startswith("OK") and compiled.startswith("OK"):
    raise SystemExit(0)

raise SystemExit(1)

Error logs

x: tensor([[ 0.,  1.,  2.,  3.],
        [ 4.,  5.,  6.,  7.],
        [ 8.,  9., 10., 11.]])
eager: IndexError select(): index 1 out of range for tensor of size [3, 1, 4] at dimension 1
compiled: OK tensor([[0.0000e+00, 1.0000e+00, 2.0000e+00, 3.0000e+00, 4.0000e+00, 5.0000e+00,
         6.0000e+00, 7.0000e+00],
        [4.0000e+00, 5.0000e+00, 6.0000e+00, 7.0000e+00, 8.0000e+00, 9.0000e+00,
         1.0000e+01, 1.1000e+01],
        [8.0000e+00, 9.0000e+00, 1.0000e+01, 1.1000e+01, 8.9683e-44, 0.0000e+00,
         1.3915e-42, 0.0000e+00]])

Versions

PyTorch version: 2.13.0.dev20260513+cpu Is debug build: False CUDA used to build PyTorch: Could not collect ROCM used to build PyTorch: N/A

OS: Ubuntu 24.04.3 LTS (x86_64) GCC version: (Ubuntu 13.3.0-6ubuntu2~24.04.1) 13.3.0 Clang version: 18.1.3 (1ubuntu1) CMake version: Could not collect Libc version: glibc-2.39

Python version: 3.11.15 (main, Mar 11 2026, 17:20:07) [GCC 14.3.0] (64-bit runtime) Python platform: Linux-6.17.0-20-generic-x86_64-with-glibc2.39 Is CUDA available: False CUDA runtime version: Could not collect CUDA_MODULE_LOADING set to: N/A GPU models and configuration: GPU 0: NVIDIA RTX 6000 Ada Generation GPU 1: NVIDIA RTX 6000 Ada Generation

Nvidia driver version: 570.211.01 cuDNN version: Could not collect Is XPU available: False HIP runtime version: N/A MIOpen runtime version: N/A Is XNNPACK available: True Caching allocator config: N/A

Versions of relevant libraries: [pip3] numpy==2.4.4 [pip3] torch==2.13.0.dev20260513+cpu [conda] numpy 2.4.4 pypi_0 pypi [conda] torch 2.13.0.dev20260513+cpu pypi_0 pypi

cc @chauhang @penguinwu @voznesenskym @EikanWang @jgong5 @Guobing-Chen @XiaobingSuper @zhuhaozhe @blzheng @wenzhe-nrv @jiayisunx @ipiszy @kadeng @muchulee8 @amjames @aakhundov @coconutruben @jataylo