-
Notifications
You must be signed in to change notification settings - Fork 25.5k
Description
🐛 Describe the bug
inductor/test_move_constructors_to_cuda.py::TestMoveConstructorsToCuda::test_multi_gpu FAILED [1.4059s] [ 14%]
==================================== RERUNS ====================================
__________________ TestMoveConstructorsToCuda.test_multi_gpu ___________________
Traceback (most recent call last):
File "/usr/lib/python3.12/unittest/case.py", line 58, in testPartExecutor
yield
File "/usr/lib/python3.12/unittest/case.py", line 634, in run
self._callTestMethod(testMethod)
File "/usr/lib/python3.12/unittest/case.py", line 589, in _callTestMethod
if method() is not None:
^^^^^^^^
File "/usr/local/lib/python3.12/dist-packages/torch/testing/_internal/common_utils.py", line 2983, in wrapper
method(*args, **kwargs)
File "/opt/pytorch/pytorch/test/inductor/test_move_constructors_to_cuda.py", line 103, in test_multi_gpu
self._check_fn(foo, True, inp)
File "/opt/pytorch/pytorch/test/inductor/test_move_constructors_to_cuda.py", line 31, in _check_fn
FileCheck().check("cpp_fused").run(code[0])
RuntimeError: Expected to find "cpp_fused" but did not find it
Searched string:
AOT ID: ['0_inference']
from ctypes import c_void_p, c_long, c_int
import torch
From CHECK: cpp_fused
To execute this test, run the following from the base repo dir:
python test/inductor/test_move_constructors_to_cuda.py TestMoveConstructorsToCuda.test_multi_gpu
tested on A100x2 systems
### Versions
nightly versions