pytorch

ee1c705c - Fix a segfault in eager_fallback when the first input is a TensorList

Fix a segfault in eager_fallback when the first input is a TensorList Summary: Pull a similar fix of CPUFallback from https://github.com/pytorch/pytorch/pull/60993/. Test Plan: Remove `cat` from ts_native_functions.yaml and `XLANativeFunctions::cat` from aten_xla_type.cpp, and then test with `test/cpp/build/test_ptltc --gtest_filter=AtenLtcTsTensorTest.TestCat`