GitLab

- size of buffer to iterate over during transfer
- if this size matches transfer size, then a single hot buffer will be used
  repeatedly for all operations

- some examples are crashing at runtime, not sure why yet

- set $HOME to avoid perturbing ~/ config files for the user running this script
- remove deprecated --dirty and LD_PRELOAD hacks
- remove explicit set of SPACK_ROOT in favor of variable set by spack itself
- other misc streamlining improvements

- also remove buffer allocation since we aren't using it

- accidentally left pmem persist commented out

- this gets the regression tests running under carns account

- this is a workaround to strip unusual characters out of output so that
  the mailx command will put it inline with message

idiomatic spack sets SPACK_ROOT and sources env based on that

Whoops: while I was here, build ior-mobject which will pull in just
about every one of our SDS packages

doesn't save a ton of time: mvapich2 takes much longer, but it's not a
package we need to build on cooley

- build and run against mvapich with verbs support
- use newer gcc
- don't do dirty builds

- no longer any need for fork of spack (PR merged)
- no longer any need for @develop libfabric

- back out a recent opa-psm2 change
- updated packages.yaml reuqires a specific compiler

- not completely working yet, but seems to be close

- check separately for requirements for the pmdk, margo, and bake benchmarks
- compile whichever ones are compatible with the environment
- this is meant to make it easier to run certain benchmarks in
  incomplete environments

- fastest config is 4 rpc handler threads, pipelining, on socket 2

Looks like if we set SPACK_ROOT in this script we can isolate spack from
our onw environment