[distributed] torch._dynamo.exc.Unsupported: Attempted to call function marked as skipped

[zxd1997066]

🐛 Describe the bug

error:
torch._dynamo.exc.Unsupported: Attempted to call function marked as skipped

cases

Class name	Test name
----------	---------
test.distributed._composable.fsdp.test_fully_shard_compile.TestFullyShardCompile	test_compiled_autograd_ctx
test.distributed._composable.fsdp.test_fully_shard_compile.TestFullyShardCompile	test_dynamo_recompiles_on_fsdp_layers
test.distributed._composable.fsdp.test_fully_shard_compile.TestFullyShardCompile	test_nested_fully_shard_backend_aot_eager
test.distributed._composable.fsdp.test_fully_shard_compile.TestFullyShardCompile	test_nested_fully_shard_backend_aot_eager_decomp_partition
test.distributed._composable.fsdp.test_fully_shard_compile.TestFullyShardCompile	test_simple_mlp_fullgraph_backend_aot_eager
test.distributed._composable.fsdp.test_fully_shard_compile.TestFullyShardCompile	test_simple_mlp_fullgraph_backend_aot_eager_decomp_partition
test.distributed._composable.fsdp.test_fully_shard_compile.TestFullyShardCompile	test_simple_mlp_fullgraph_backend_inductor
test.distributed._composable.fsdp.test_fully_shard_compile.TestFullyShardCompile	test_transformer_backend_aot_eager
test.distributed._composable.fsdp.test_fully_shard_compile.TestFullyShardCompile	test_transformer_backend_aot_eager_decomp_partition

FAILED [37.4646s] ../../../../test/distributed/_composable/fsdp/test_fully_shard_compile.py::TestFullyShardCompile::test_compiled_autograd_ctx - RuntimeError: Process 1 exited with error code 10 and exception:
Traceback (most recent call last):
  File "/home/sdp/.conda/envs/2025_ww17/lib/python3.10/site-packages/torch/testing/_internal/common_distributed.py", line 761, in run_test
    getattr(self, test_name)()
  File "/home/sdp/.conda/envs/2025_ww17/lib/python3.10/site-packages/torch/testing/_internal/common_distributed.py", line 634, in wrapper
    fn()
  File "/home/sdp/.conda/envs/2025_ww17/lib/python3.10/site-packages/torch/testing/_internal/common_utils.py", line 3155, in wrapper
    method(*args, **kwargs)
  File "/home/sdp/.conda/envs/2025_ww17/lib/python3.10/site-packages/torch/testing/_internal/common_utils.py", line 1876, in wrapper
    return fn(*args, **kwargs)
  File "/home/sdp/jenkins/workspace/distributed_ut_regular/pytorch/test/distributed/_composable/fsdp/test_fully_shard_compile.py", line 485, in test_compiled_autograd_ctx
    out.sum().backward()
  File "/home/sdp/.conda/envs/2025_ww17/lib/python3.10/site-packages/torch/_tensor.py", line 648, in backward
    torch.autograd.backward(
  File "/home/sdp/.conda/envs/2025_ww17/lib/python3.10/site-packages/torch/autograd/__init__.py", line 354, in backward
    _engine_run_backward(
  File "/home/sdp/.conda/envs/2025_ww17/lib/python3.10/site-packages/torch/autograd/graph.py", line 824, in _engine_run_backward
    return Variable._execution_engine.run_backward(  # Calls into the C++ engine to run the backward pass
  File "/home/sdp/.conda/envs/2025_ww17/lib/python3.10/site-packages/torch/_dynamo/compiled_autograd.py", line 963, in runtime_wrapper
    out = compiled_fn(inputs, sizes, scalars, hooks, packed_inputs)
  File "/home/sdp/.conda/envs/2025_ww17/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
    return self._call_impl(*args, **kwargs)
  File "/home/sdp/.conda/envs/2025_ww17/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
    return forward_call(*args, **kwargs)
  File "/home/sdp/.conda/envs/2025_ww17/lib/python3.10/site-packages/torch/_dynamo/eval_frame.py", line 671, in _fn
    raise e.with_traceback(None) from e.__cause__
torch._dynamo.exc.Unsupported: Attempted to call function marked as skipped
  Explanation: Dynamo developers have intentionally marked that the function `current_stream` in file `/home/sdp/.conda/envs/2025_ww17/lib/python3.10/site-packages/torch/xpu/__init__.py` should not be traced.
  Hint: Avoid calling the function `current_stream`.
  Hint: Remove the function `current_stream` or the file `/home/sdp/.conda/envs/2025_ww17/lib/python3.10/site-packages/torch/xpu/__init__.py` from torch/_dynamo/trace_rules.py. More graph breaks may occur as a result of attempting to trace into the function.
  Hint: Please file an issue to PyTorch.

  Developer debug context: module: torch.xpu, qualname: current_stream, skip reason: <missing reason>


from user code:
   File "<eval_with_key>.0", line 45, in forward
    _exec_final_callbacks_stub = torch__dynamo_external_utils__exec_final_callbacks_stub();  _exec_final_callbacks_stub = None
  File "/home/sdp/.conda/envs/2025_ww17/lib/python3.10/site-packages/torch/_dynamo/external_utils.py", line 165, in exec_final_callbacks
    cb()
  File "/home/sdp/.conda/envs/2025_ww17/lib/python3.10/site-packages/torch/distributed/fsdp/_fully_shard/_fsdp_state.py", line 297, in _root_post_backward_final_callback
    fsdp_param_group.post_backward()
  File "/home/sdp/.conda/envs/2025_ww17/lib/python3.10/site-packages/torch/distributed/fsdp/_fully_shard/_fsdp_param_group.py", line 445, in post_backward
    ) = foreach_reduce(
  File "/home/sdp/.conda/envs/2025_ww17/lib/python3.10/site-packages/torch/utils/_contextlib.py", line 116, in decorate_context
    return func(*args, **kwargs)
  File "/home/sdp/.conda/envs/2025_ww17/lib/python3.10/site-packages/torch/distributed/fsdp/_fully_shard/_fsdp_collectives.py", line 406, in foreach_reduce
    current_stream = device_handle.current_stream()

Set TORCHDYNAMO_VERBOSE=1 for the internal stack trace (please do this especially if you're reporting a bug to PyTorch). For even more developer context, set TORCH_LOGS="+dynamo"


To execute this test, run the following from the base repo dir:
    PYTORCH_TEST_WITH_SLOW=1 python test/distributed/_composable/fsdp/test_fully_shard_compile.py TestFullyShardCompile.test_compiled_autograd_ctx

Versions

PyTorch version: 2.8.0a0+gite558eaa
Is debug build: False
CUDA used to build PyTorch: None
ROCM used to build PyTorch: N/A

OS: Ubuntu 22.04.5 LTS (x86_64)
GCC version: (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0
Clang version: Could not collect
CMake version: version 4.0.0
Libc version: glibc-2.35

Python version: 3.10.17 | packaged by conda-forge | (main, Apr 10 2025, 22:19:12) [GCC 13.3.0] (64-bit runtime)
Python platform: Linux-5.15.0-131-generic-x86_64-with-glibc2.35
Is CUDA available: False
CUDA runtime version: No CUDA
CUDA_MODULE_LOADING set to: N/A
GPU models and configuration: No CUDA
Nvidia driver version: No CUDA
cuDNN version: No CUDA
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, 57 bits virtual
Byte Order: Little Endian
CPU(s): 192
On-line CPU(s) list: 0-191
Vendor ID: GenuineIntel
Model name: Intel(R) Xeon(R) Platinum 8468V
CPU family: 6
Model: 143
Thread(s) per core: 2
Core(s) per socket: 48
Socket(s): 2
Stepping: 8
Frequency boost: enabled
CPU max MHz: 2401.0000
CPU min MHz: 800.0000
BogoMIPS: 4800.00
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 tsc_known_freq 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 cat_l2 cdp_l3 invpcid_single intel_ppin cdp_l2 ssbd mba ibrs ibpb stibp ibrs_enhanced tpr_shadow vnmi flexpriority ept vpid ept_ad fsgsbase tsc_adjust bmi1 avx2 smep bmi2 erms invpcid cqm rdt_a avx512f avx512dq rdseed adx smap avx512ifma clflushopt clwb intel_pt avx512cd sha_ni avx512bw avx512vl xsaveopt xsavec xgetbv1 xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local split_lock_detect avx_vnni avx512_bf16 wbnoinvd dtherm ida arat pln pts avx512vbmi umip pku ospke waitpkg avx512_vbmi2 gfni vaes vpclmulqdq avx512_vnni avx512_bitalg tme avx512_vpopcntdq la57 rdpid bus_lock_detect cldemote movdiri movdir64b enqcmd fsrm md_clear serialize tsxldtrk pconfig arch_lbr amx_bf16 avx512_fp16 amx_tile amx_int8 flush_l1d arch_capabilities
Virtualization: VT-x
L1d cache: 4.5 MiB (96 instances)
L1i cache: 3 MiB (96 instances)
L2 cache: 192 MiB (96 instances)
L3 cache: 195 MiB (2 instances)
NUMA node(s): 2
NUMA node0 CPU(s): 0-47,96-143
NUMA node1 CPU(s): 48-95,144-191
Vulnerability Gather data sampling: Not affected
Vulnerability Itlb multihit: Not affected
Vulnerability L1tf: Not affected
Vulnerability Mds: Not affected
Vulnerability Meltdown: Not affected
Vulnerability Mmio stale data: Not affected
Vulnerability Reg file data sampling: Not affected
Vulnerability Retbleed: Not affected
Vulnerability Spec rstack overflow: Not affected
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl and seccomp
Vulnerability Spectre v1: Mitigation; usercopy/swapgs barriers and __user pointer sanitization
Vulnerability Spectre v2: Mitigation; Enhanced / Automatic IBRS; IBPB conditional; RSB filling; PBRSB-eIBRS SW sequence; BHI BHI_DIS_S
Vulnerability Srbds: Not affected
Vulnerability Tsx async abort: Not affected

Versions of relevant libraries:
[pip3] numpy==1.26.4
[pip3] optree==0.15.0
[pip3] torch==2.8.0a0+gite558eaa
[pip3] torchao==0.11.0+git37d161e8
[pip3] torchdata==0.11.0
[pip3] torchtune==0.0.0
[pip3] torchvision==0.22.0a0+5f03dc5
[pip3] triton==3.3.0+git85788e6d
[conda] numpy 1.26.4 pypi_0 pypi
[conda] optree 0.15.0 pypi_0 pypi
[conda] torch 2.8.0a0+gite558eaa pypi_0 pypi
[conda] torchao 0.11.0+git37d161e8 dev_0
[conda] torchdata 0.11.0 pypi_0 pypi
[conda] torchtune 0.0.0 pypi_0 pypi
[conda] torchvision 0.22.0a0+5f03dc5 pypi_0 pypi
[conda] triton 3.3.0+git85788e6d pypi_0 pypi

[daisyden]

@guangyey could you help to check whether current_stream() need to be skipped when attempting to trace?

[guangyey]

@daisyden OK I will take a look.

[guangyey]

@zxd1997066 @daisyden could you help to try if this case could pass if adding torch.xpu.current_stream into torch_non_c_binding_in_graph_functions mentioned in https://github.com/pytorch/pytorch/blob/5c01302cc812e9fc4282a09da5655f0ee57d9c5c/torch/_dynamo/trace_rules.py#L2261

[daisyden]

We need also to add torch.xpu.stream int this list. With both torch.xpu.stream and torch.xpu.current_stream added, the case test_compiled_autograd_ctx passed.

[daisyden]

Other cases are verified with patch, only the two need to revisit after timeout issue introduced by 7ee006f is fixed.
test_simple_mlp_fullgraph_backend_inductor
test_compiled_autograd_ctx

diff --git a/torch/_dynamo/trace_rules.py b/torch/_dynamo/trace_rules.py
index f04e360894..1f8430ca2c 100644
--- a/torch/_dynamo/trace_rules.py
+++ b/torch/_dynamo/trace_rules.py
@@ -2260,6 +2260,10 @@ if sys.version_info >= (3, 11):
 # or captures global state, please add it to manual_torch_name_rule_map instead
 torch_non_c_binding_in_graph_functions = dict.fromkeys(
     [
+        "torch.xpu.current_stream",
+        "torch.xpu.stream",
+        "torch.xpu.Stream",
+        "torch.xpu.Event",
         "torch.__future__.get_overwrite_module_params_on_conversion",
         "torch.__future__.set_overwrite_module_params_on_conversion",
         "torch.__getattr__",
diff --git a/torch/_dynamo/variables/user_defined.py b/torch/_dynamo/variables/user_defined.py
index e10c5ae62e..d0111ea161 100644
--- a/torch/_dynamo/variables/user_defined.py
+++ b/torch/_dynamo/variables/user_defined.py
@@ -160,6 +160,8 @@ class UserDefinedClassVariable(UserDefinedVariable):
             torch.Tensor,
             torch.cuda.Stream,
             torch.cuda.Event,
+            torch.xpu.Stream,
+            torch.xpu.Event,
         }
         if hasattr(torch, "hpu"):
             _in_graph_class_list.update(

[guangyey]

torch.xpu.Stream and torch.xpu.Event should not be included in torch_non_c_binding_in_graph_functions

[daisyden]

@guangyey You are right, with the update all the cases passed.