Magically pick up Slurm CPU limits via SLURM_JOB_CPUS_PER_NODE

[adamnovak]

Changelog Entry

To be copied to the draft changelog by merger:

• vg now tries to limit itself to a good number of threads for the number of CPUs in any enclosing Slurm job, via SLURM_JOB_CPUS_PER_NODE and CPU affinity masks.

Description

This should prevent vg from using hundreds of threads on small Slurm jobs run on large machines, without the user needing to remember to manually pass -t in their Slurm script.