We are having what appears to be the same segfault issue when running some of our larger chained functions (likely with quite large plans). However, we can reliably trigger this by looping over `withColumns` to increase the plan size. Our case is also a segfault and ends up in the protobuf library as well. Has anyone found success with increasing `spark.connect.grpc.maxInboundMessageSize`, besides refactoring to flatten the plan as much as possible.