Open
Description
🐛 Describe the bug
I am trying to use torch compile on my functions and encounter this issue. I attached a minimum test program so anyone can reproduce the issue.
from dataclasses import dataclass
import torch
@dataclass(frozen=True)
class BaseFlexData:
dtype: torch.dtype | None = None
def view(self, x: torch.Tensor):
if self.dtype is None:
return x
return x.view(self.dtype)
def reinterpret(self, x):
if self.dtype is None or x.dtype.itemsize > 1:
return x
return x.view(self.dtype)
@dataclass(frozen=True)
class InFlexData(BaseFlexData):
scale: torch.Tensor | None = None
@property
def is_per_batch(self):
return False if self.scale is None else len(self.scale) > 1
@dataclass(frozen=True)
class OutFlexData(BaseFlexData):
expected_scale: torch.Tensor | None = None
actual_scale: torch.Tensor | None = None
checksum_scale: torch.Tensor | None = None
def __iter__(self):
yield self<
8000
/span>.expected_scale
yield self.actual_scale
yield self.checksum_scale
@dataclass(frozen=True)
class FlexCtx:
lhs_data: InFlexData = InFlexData()
rhs_data: InFlexData = InFlexData()
out_data: OutFlexData = OutFlexData()
@dataclass
class DummyClass:
flex_ctx: FlexCtx = FlexCtx()
def __post_init__(self):
assert self.flex_ctx.rhs_data.scale is None, "flex and mx_ctx cannot be used together"
@torch.compile(fullgraph=True)
def dummy_method():
var = DummyClass(flex_ctx=FlexCtx(rhs_data=InFlexData()))
return var
dummy_method()Error logs
TORCHDYNAMO_VERBOSE=1 python test_compile.py
Traceback (most recent call last):
File "/home/eecs/yongye.zhu/vllm/tests/kernels/moe/test_compile.py", line 56, in <module>
dummy_method()
File "/home/eecs/yongye.zhu/miniconda3/envs/vllm/lib/python3.12/site-packages/torch/_dynamo/eval_frame.py", line 685, in _fn
return fn(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^
File "/home/eecs/yongye.zhu/miniconda3/envs/vllm/lib/python3.12/site-packages/torch/_dynamo/convert_frame.py", line 1463, in __call__
return self._torchdynamo_orig_callable(
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/eecs/yongye.zhu/miniconda3/envs/vllm/lib/python3.12/site-packages/torch/_dynamo/convert_frame.py", line 624, in __call__
return _compile(
^^^^^^^^^
File "/home/eecs/yongye.zhu/miniconda3/envs/vllm/lib/python3.12/site-packages/torch/_dynamo/convert_frame.py", line 1087, in _compile
guarded_code = compile_inner(code, one_graph, hooks, transform)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/eecs/yongye.zhu/miniconda3/envs/vllm/lib/python3.12/site-packages/torch/_utils_internal.py", line 97, in wrapper_function
return function(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/eecs/yongye.zhu/miniconda3/envs/vllm/lib/python3.12/site-packages/torch/_dynamo/convert_frame.py", line 778, in compile_inner
return _compile_inner(code, one_graph, hooks, transform)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/eecs/yongye.zhu/miniconda3/envs/vllm/lib/python3.12/site-packages/torch/_dynamo/convert_frame.py", line 817, in _compile_inner
out_code = transform_code_object(code, transform)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/eecs/yongye.zhu/miniconda3/envs/vllm/lib/python3.12/site-packages/torch/_dynamo/bytecode_transformation.py", line 1423, in transform_code_object
transformations(instructions, code_options)
File "/home/eecs/yongye.zhu/miniconda3/envs/vllm/lib/python3.12/site-packages/torch/_dynamo/convert_frame.py", line 264, in _fn
return fn(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^
File "/home/eecs/yongye.zhu/miniconda3/envs/vllm/lib/python3.12/site-packages/torch/_dynamo/convert_frame.py", line 742, in transform
tracer.run()
File "/home/eecs/yongye.zhu/miniconda3/envs/vllm/lib/python3.12/site-packages/torch/_dynamo/symbolic_convert.py", line 3508, in run
super().run()
File "/home/eecs/yongye.zhu/miniconda3/envs/vllm/lib/python3.12/site-packages/torch/_dynamo/symbolic_convert.py", line 1345, in run
while self.step():
^^^^^^^^^^^
File "/home/eecs/yongye.zhu/miniconda3/envs/vllm/lib/python3.12/site-packages/torch/_dynamo/symbolic_convert.py", line 1253, in step
self.dispatch_table[inst.opcode](self, inst)
File "/home/eecs/yongye.zhu/miniconda3/envs/vllm/lib/python3.12/site-packages/torch/_dynamo/symbolic_convert.py", line 828, in wrapper
return inner_fn(self, inst)
^^^^^^^^^^^^^^^^^^^^
File "/home/eecs/yongye.zhu/miniconda3/envs/vllm/lib/python3.12/site-packages/torch/_dynamo/symbolic_convert.py", line 2934, in CALL
self._call(inst)
File "/home/eecs/yongye.zhu/miniconda3/envs/vllm/lib/python3.12/site-packages/torch/_dynamo/symbolic_convert.py", line 2928, in _call
self.call_function(fn, args, kwargs)
File "/home/eecs/yongye.zhu/miniconda3/envs/vllm/lib/python3.12/site-packages/torch/_dynamo/symbolic_convert.py", line 1179, in call_function
self.push(fn.call_function(self, args, kwargs)) # type: ignore[arg-type]
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/eecs/yongye.zhu/miniconda3/envs/vllm/lib/python3.12/site-packages/torch/_dynamo/variables/lazy.py", line 201, in realize_and_forward
return getattr(self.realize(), name)(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/eecs/yongye.zhu/miniconda3/envs/vllm/lib/python3.12/site-packages/torch/_dynamo/variables/user_defined.py", line 603, in call_function
var_tracker = VariableTracker.build(tx, field.default)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/eecs/yongye.zhu/miniconda3/envs/vllm/lib/python3.12/site-packages/torch/_dynamo/variables/base.py", line 592, in build
return builder.SourcelessBuilder.create(tx, value)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/eecs/yongye.zhu/miniconda3/envs/vllm/lib/python3.12/site-packages/torch/_dynamo/variables/builder.py", line 3387, in create
unimplemented_v2(
File "/home/eecs/yongye.zhu/miniconda3/envs/vllm/lib/python3.12/site-packages/torch/_dynamo/exc.py", line 517, in unimplemented_v2
raise Unsupported(msg)
torch._dynamo.exc.Unsupported: Unexpected type in sourceless builder
Explanation: SourcelessBuilder.create does not know how to wrap <class '__main__.InFlexData'>
Hint: This is likely to be a Dynamo bug. Please report an issue to PyTorch.
Developer debug context: __main__.InFlexData
from user code:
File "/home/eecs/yongye.zhu/vllm/tests/kernels/moe/test_compile.py", line 53, in dummy_method
var = DummyClass(flex_ctx=FlexCtx(rhs_data=InFlexData()))
Versions
Collecting environment information...
PyTorch version: 2.8.0.dev20250516+cu126
Is debug build: False
CUDA used to build PyTorch: 12.6
ROCM used to build PyTorch: N/A
OS: Ubuntu 24.04.2 LTS (x86_64)
GCC version: (Ubuntu 13.3.0-6ubuntu2~24.04) 13.3.0
Clang version: 18.1.3 (1ubuntu1)
CMake version: version 3.31.6
Libc version: glibc-2.39
Python version: 3.12.9 | packaged by Anaconda, Inc. | (main, Feb 6 2025, 18:56:27) [GCC 11.2.0] (64-bit runtime)
Python platform: Linux-6.8.0-60-generic-x86_64-with-glibc2.39
Is CUDA available: True
CUDA runtime version: 12.6.85
CUDA_MODULE_LOADING set to: LAZY
GPU models and configuration:
GPU 0: NVIDIA RTX A6000
GPU 1: NVIDIA RTX A6000
GPU 2: NVIDIA RTX A6000
GPU 3: NVIDIA RTX A6000
GPU 4: NVIDIA RTX A6000
GPU 5: NVIDIA RTX A6000
GPU 6: NVIDIA RTX A6000
GPU 7: NVIDIA RTX A6000
Nvidia driver version: 560.35.05
cuDNN version: Probably one of the following:
/usr/local/cuda-11.3/targets/x86_64-linux/lib/libcudnn.so.8.2.0
/usr/local/cuda-11.3/targets/x86_64-linux/lib/libcudnn_adv_infer.so.8.2.0
/usr/local/cuda-11.3/targets/x86_64-linux/lib/libcudnn_adv_train.so.8.2.0
/usr/local/cuda-11.3/targets/x86_64-linux/lib/libcudnn_cnn_infer.so.8.2.0
/usr/local/cuda-11.3/targets/x86_64-linux/lib/libcudnn_cnn_train.so.8.2.0
/usr/local/cuda-11.3/targets/x86_64-linux/lib/libcudnn_ops_infer.so.8.2.0
/usr/local/cuda-11.3/targets/x86_64-linux/lib/libcudnn_ops_train.so.8.2.0
/usr/local/cuda-12.6/targets/x86_64-linux/lib/libcudnn.so.9.5.1
/usr/local/cuda-12.6/targets/x86_64-linux/lib/libcudnn_adv.so.9.5.1
/usr/local/cuda-12.6/targets/x86_64-linux/lib/libcudnn_cnn.so.9.5.1
/usr/local/cuda-12.6/targets/x86_64-linux/lib/libcudnn_engines_precompiled.so.9.5.1
/usr/local/cuda-12.6/targets/x86_64-linux/lib/libcudnn_engines_runtime_compiled.so.9.5.1
/usr/local/cuda-12.6/targets/x86_64-linux/lib/libcudnn_graph.so.9.5.1
/usr/local/cuda-12.6/targets/x86_64-linux/lib/libcudnn_heuristic.so.9.5.1
/usr/local/cuda-12.6/targets/x86_64-linux/lib/libcudnn_ops.so.9.5.1
HIP runtime version: N/A
MIOpen runtime version: N/A
Is XNNPACK available: True
CPU:
Architecture: x86_64
CPU op-mode(s): 32-bit, 64-bit
Address sizes: 46 bits physical, 48 bits virtual
Byte Order: Little Endian
CPU(s): 48
On-line CPU(s) list: 0-47
Vendor ID: GenuineIntel
Model name: Intel(R) Xeon(R) Gold 6126 CPU @ 2.60GHz
CPU family: 6
Model: 85
Thread(s) per core: 2
Core(s) per socket: 12
Socket(s): 2
Stepping: 4
CPU(s) scaling MHz: 40%
CPU max MHz: 3700.0000
CPU min MHz: 1000.0000
BogoMIPS: 5200.00
Flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx pdpe1gb rdtscp lm constant_tsc art arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc cpuid aperfmperf pni pclmulqdq dtes64 monitor ds_cpl vmx smx est tm2 ssse3 sdbg fma cx16 xtpr pdcm pcid dca sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes xsave avx f16c rdrand lahf_lm abm 3dnowprefetch cpuid_fault epb cat_l3 cdp_l3 pti intel_ppin ssbd mba ibrs ibpb stibp tpr_shadow flexpriority ept vpid ept_ad fsgsbase tsc_adjust bmi1 hle avx2 smep bmi2 erms invpcid rtm cqm mpx rdt_a avx512f avx512dq rdseed adx smap clflushopt clwb intel_pt avx512cd avx512bw avx512vl xsaveopt xsavec xgetbv1 xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local dtherm ida arat pln pts hwp hwp_act_window hwp_epp hwp_pkg_req vnmi pku ospke md_clear flush_l1d arch_capabilities
Virtualization: VT-x
L1d cache: 768 KiB (24 instances)
L1i cache: 768 KiB (24 instances)
L2 cache: 24 MiB (24 instances)
L3 cache: 38.5 MiB (2 instances)
NUMA node(s): 2
NUMA node0 CPU(s): 0-11,24-35
NUMA node1 CPU(s): 12-23,36-47
Vulnerability Gather data sampling: Mitigation; Microcode
Vulnerability Itlb multihit: KVM: Mitigation: VMX disabled
Vulnerability L1tf: Mitigation; PTE Inversion; VMX conditional cache flushes, SMT vulnerable
Vulnerability Mds: Mitigation; Clear CPU buffers; SMT vulnerable
Vulnerability Meltdown: Mitigation; PTI
Vulnerability Mmio stale data: Mitigation; Clear CPU buffers; SMT vulnerable
Vulnerability Reg file data sampling: Not affected
Vulnerability Retbleed: Mitigation; IBRS
Vulnerability Spec rstack overflow: Not affected
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1: Mitigation; usercopy/swapgs barriers and __user pointer sanitization
Vulnerability Spectre v2: Mitigation; IBRS; IBPB conditional; STIBP conditional; RSB filling; PBRSB-eIBRS Not affected; BHI Not affected
Vulnerability Srbds: Not affected
Vulnerability Tsx async abort: Mitigation; Clear CPU buffers; SMT vulnerable
Versions of relevant libraries:
[pip3] mypy-extensions==1.0.0
[pip3] numpy==1.26.4
[pip3] nvidia-cublas-cu12==12.6.4.1
[pip3] nvidia-cuda-cupti-cu12==12.6.80
[pip3] nvidia-cuda-nvrtc-cu12==12.6.77
[pip3] nvidia-cuda-runtime-cu12==12.6.77
[pip3] nvidia-cudnn-cu12==9.5.1.17
[pip3] nvidia-cufft-cu12==11.3.0.4
[pip3] nvidia-curand-cu12==10.3.7.77
[pip3] nvidia-cusolver-cu12==11.7.1.2
[pip3] nvidia-cusparse-cu12==12.5.4.2
[pip3] nvidia-cusparselt-cu12==0.6.3
[pip3] nvidia-nccl-cu12==2.26.5
[pip3] nvidia-nvjitlink-cu12==12.6.85
[pip3] nvidia-nvtx-cu12==12.6.77
[pip3] pytorch-triton==3.3.0+git96316ce5
[pip3] torch==2.8.0.dev20250516+cu126
[pip3] torchaudio==2.6.0.dev20250516+cu126
[pip3] torchvision==0.22.0.dev20250516+cu126
[pip3] triton==3.3.0+gite6b9efdf
[pip3] triton_kernels==1.0.0
[pip3] tritonclient==2.51.0
[pip3] vector-quantize-pytorch==1.21.2
[conda] numpy 1.26.4 pypi_0 pypi
[conda] nvidia-cublas-cu12 12.6.4.1 pypi_0 pypi
[conda] nvidia-cuda-cupti-cu12 12.6.80 pypi_0 pypi
[conda] nvidia-cuda-nvrtc-cu12 12.6.77 pypi_0 pypi
[conda] nvidia-cuda-runtime-cu12 12.6.77 pypi_0 pypi
[conda] nvidia-cudnn-cu12 9.5.1.17 pypi_0 pypi
[conda] nvidia-cufft-cu12 11.3.0.4 pypi_0 pypi
[conda] nvidia-curand-cu12 10.3.7.77 pypi_0 pypi
[conda] nvidia-cusolver-cu12 11.7.1.2 pypi_0 pypi
[conda] nvidia-cusparse-cu12 12.5.4.2 pypi_0 pypi
[conda] nvidia-cusparselt-cu12 0.6.3 pypi_0 pypi
[conda] nvidia-nccl-cu12 2.26.5 pypi_0 pypi
[conda] nvidia-nvjitlink-cu12 12.6.85 pypi_0 pypi
[conda] nvidia-nvtx-cu12 12.6.77 pypi_0 pypi
[conda] pytorch-triton 3.3.0+git96316ce5 pypi_0 pypi
[conda] torch 2.8.0.dev20250516+cu126 pypi_0 pypi
[conda] torchaudio 2.6.0.dev20250516+cu126 pypi_0 pypi
[conda] torchvision 0.22.0.dev20250516+cu126 pypi_0 pypi
[conda] triton 3.3.0+gite6b9efdf pypi_0 pypi
[conda] triton-kernels 1.0.0 pypi_0 pypi
[conda] tritonclient 2.51.0 pypi_0 pypi
[conda] vector-quantize-pytorch 1.21.2 pypi_0 pypi
cc @chauhang @penguinwu @voznesenskym @EikanWang @jgong5 @Guobing-Chen @XiaobingSuper @zhuhaozhe @blzheng @wenzhe-nrv @jiayisunx @chenyang78 @kadeng @amjames