Re-raise in _run_sympy_handler to reduce log spew (#138356)

bobrenjc93 · pytorchmergebot · commit 38ea4873386e · 2024-10-19T16:02:45.000Z
Fixes: #138069 I tested this by running `python test/inductor/test_torchinductor_dynamic_shapes.py DynamicShapesCpuTests.test_builtins_round_float_ndigits_pos_dynamic_shapes_cpu` before and after the change and verifying no more log spew. I'm uncertain on if it makes sense to add a test for this PR. Question for reviewers: is there a standard paradigm for testing these log spew based fixed? Happy to add a test if someone can point me towards the right direction. Pull Request resolved: #138356 Approved by: https://github.com/ezyang
diff --git a/torch/utils/_sympy/interp.py b/torch/utils/_sympy/interp.py
@@ -155,6 +155,8 @@ def _run_sympy_handler(analysis, args, expr, index_dtype=torch.int64):
             r = handler(*args)
             log.debug("%s(%s) -> %s", handler_name, args, r)
             return r
+    except NotImplementedError:
+        raise
     except Exception:
         log.warning("failed while executing %s(%s)", handler_name, args)
         raise
