LockFreeImmortalSpace is failing in CI

[qinsoon]

https://www.mmtk.io/ci-perf-result/master/mutator_history.html shows the results for both nogc plans using LockFreeImmortalSpace are 0. In the log, it panics with

thread '<unnamed>' panicked at 'Out of virtual address space at 0x40000000 (0x1f400040000000 > 0x800000000)', /home/gitlab-runner/actions-runner/_work/mmtk-core/mmtk-core/mmtk-openjdk/repos/mmtk-core/src/util/heap/heap_meta.rs:27:9
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace
fatal runtime error: failed to initiate panic, error 5

I guess it might be related with compressed pointer.

---

LockFreeImmortalSpace was not running properly in our CI. It was built with LockFreeImmortalSpace, but was actually running the default plan (probably since the beginning), which was a mistake in the CI script. There were also some issues with LockFreeImmortalSpace so it did not actually run. but the issues were identified and fixed in #1003. When we moved to running-ng in our CI (#1005), the run configs for nogc plans are fixed. So those plans with LockFreeImmortalSpace are benchmarked properly, but the runs failed with the error message above.