[DTensor] Replaced neg dim normalization with assert in helper (#114141)

This is a replacement for #113922. I think we can still leave the check for negative shard dimension in `compute_local_shape_and_global_offset` and replace the normalization logic with an assert. This should provide us a stack trace to see which user-facing API did not normalize the dim as expected.
Pull Request resolved: #114141
Approved by: https://github.com/wanchaol
ghstack dependencies: #113919, #113924, #114134, #113925, #113930

torch/distributed/_tensor/_utils.py

@@ -145,8 +145,10 @@ def compute_global_tensor_info(
         if placement.is_shard():
             shard_placement = cast(Shard, placement)
             if shard_placement.dim < 0:
-                # normalize shard dim to be positive
-                shard_placement.dim += len(tensor_shape)
+                raise AssertionError(
+                    "Shard placements should have negative dims normalized in "
+                    f"the user-facing APIs: {shard_placement}"
+                )
             shard_dim = shard_placement.dim
 
             assert (