Validate tensor sizes during method load (#7663)

Summary:

If a tensor's `sizes` array contains a negative value, the TensorImpl ctor will panic.

Validate sizes before creating the TensorImpl to treat this as a non-fatal error when loading malformed PTE files.

Reviewed By: JacobSzwejbka

Differential Revision: D68180029

Author: dbort

runtime/executor/tensor_parser_portable.cpp

@@ -101,6 +101,19 @@ Result<Tensor> parseTensor(
     sizes = const_cast<exec_aten::SizesType*>(serialized_sizes);
     dim_order = const_cast<exec_aten::DimOrderType*>(serialized_dim_order);
   }
+  // Validate sizes before using them in case the PTE data is bad. We can't
+  // detect bad positive values, but we can reject negative values, which would
+  // otherwise panic in the TensorImpl ctor. dim_order_to_stride() will validate
+  // dim_order.
+  for (int i = 0; i < dim; i++) {
+    ET_CHECK_OR_RETURN_ERROR(
+        sizes[i] >= 0,
+        InvalidProgram,
+        "Negative size[%d] %" PRId32,
+        i,
+        sizes[i]);
+  }
+
   // We will remove strides from schema.
   // Allocating strides buffer here and populating it.
   // In subsequent diffs we can remove strides accessor, however this