ruby --version Segmentation fault on NVIDIA Grace

The results below were the same with and without using eessi_container.sh:

Running ruby --version seg faults on NVIDIA Grace CPU node in site A:

{EESSI 2023.06} user@j21n14:~$ /cvmfs/software.eessi.io/versions/2023.06/software/linux/aarch64/nvidia/grace/software/Ruby/3.3.0-GCCcore-12.3.0/bin/ruby --version
Segmentation fault (core dumped)

Running ruby --version from neoverse_v1 stack:

{EESSI 2023.06} user@j21n14:~$ /cvmfs/software.eessi.io/versions/2023.06/software/linux/aarch64/neoverse_v1/software/Ruby/3.3.0-GCCcore-12.3.0/bin/ruby --version
Segmentation fault (core dumped)

While on another NVIDIA Grace CPU node in site B:

{EESSI 2023.06} [user@jrc0900 ~]$ /cvmfs/software.eessi.io/versions/2023.06/software/linux/aarch64/nvidia/grace/software/Ruby/3.3.0-GCCcore-12.3.0/bin/ruby  --version
ruby 3.3.0 (2023-12-25 revision 5124f9ac75) [aarch64-linux]

Running ruby --version from neoverse_v1 stack:

{EESSI 2023.06} [user@jrc0900 ~]$ /cvmfs/software.eessi.io/versions/2023.06/software/linux/aarch64/neoverse_v1/software/Ruby/3.3.0-GCCcore-12.3.0/bin/ruby  --version
ruby 3.3.0 (2023-12-25 revision 5124f9ac75) [aarch64-linux]